Flexible biomimetic muscle complex, control method thereof, driving module and robot

Abstract

This application discloses a flexible biomimetic muscle complex and its control method, drive module, and robot, belonging to the field of biomimetic robot technology. The complex includes: multiple strip-shaped components, whose deformation behavior is anisotropic at least under electric field excitation; it includes a first component group and a second component group, with the axial stiffness of the first component group being higher than that of the second component group; the first and second component groups are coupled through a preset force transmission structure. When excited by the electric field, the first driving force generated by the first component group, including a first deformation, and the second driving force generated by the second component group, including a second deformation, cooperate spatially to achieve motion output consistent with the characteristics of biological muscle movement. This application simulates the cooperative working mechanism of fast and slow muscle fibers in biological muscles by preset component groups with different moduli, achieving endogenous synergy of contraction and bending without the need for complex control algorithms, reducing control complexity, and improving biomimetic performance.

Description

Technical Field

[0001] This application relates to the field of biomimetic robot technology, specifically to a flexible biomimetic muscle complex and its control method, drive module, and robot. Background Technology

[0002] Bionic actuation technology is the core foundation for improving the anthropomorphic performance and naturalness of robot interactions. An ideal bionic actuation system should possess excellent compliance, high energy density, rapid response, and quiet operation, especially in fields requiring precise motion control such as robot facial expressions, bionic dexterous hands, and wearable devices, where these properties are urgently needed. In recent years, significant progress has been made in research on flexible electro-actuated artificial muscles, mainly forming the following technical paths:

[0003] The first category is based on the intrinsic performance enhancement path of disordered materials. Electroactuated materials can undergo active deformation under electric field excitation, which is the core foundation for realizing flexible actuation. Early research focused on isotropic disordered materials, significantly improving the electromechanical sensitivity and energy density of the materials through molecular structure design, cross-linking network regulation, or functional filler composites. This type of research has broken through the bottleneck of actuation performance at the material level, but because the materials themselves lack directional structure, the deformation obtained is usually isotropic in-plane expansion, which still requires external constraints or complex control to be transformed into directional contraction, bending, and other mechanical motion outputs.

[0004] The second category is deformation-guided paths based on intrinsically anisotropic materials. To address the problem of uncontrollable deformation directions in disordered materials, researchers have attempted to endow them with anisotropy from the material's origin. Typical approaches include using molecular orientation, mechanical and thermal training, and directional filler arrangement to give the material itself anisotropic mechanical properties—exhibiting different Young's moduli or driving responses in different directions. These intrinsically anisotropic materials can directly generate directional deformation under electric field excitation, such as contraction or elongation along the molecular orientation direction. However, such approaches are still limited to the scope of "homogenized units": the material properties of a single fiber or a single-layer film are uniform, and its deformation mode is determined by the material itself, making it difficult to achieve a combined function of contraction and bending within the same unit. If combined motion is required, multiple homogeneous fibers still need to be simply bundled together, and the switching of motion modes is achieved through independent addressing control of individual fibers. Essentially, this belongs to "intelligent selection" at the control level rather than "endogenous synergy" at the structural level.

[0005] The third category is deformation control based on external constraints. Another approach involves imposing constraints on disordered materials through external structures, rather than modifying the materials themselves anisotropically, guiding their isotropic deformation into directional output. Typical solutions include: integrating the constraint-guiding component with the actuator, such as adding fiber reinforcement layers, mesh constraint layers, or discrete constraint elements to the surface of the actuator; or constructing an external three-dimensional constraint structure independent of the actuator, using its pre-defined spatial non-uniform stiffness distribution to guide and transform the actuator's volume expansion into macroscopic mechanical motion. This type of solution achieves directional control of deformation through "external constraints," effectively improving the controllability and output efficiency of flexible actuators. However, the constraint and actuator functions are strongly coupled, and once the motion pattern is fixed, it is difficult to reconstruct, and it remains at the level of guiding the deformation of a single material unit.

[0006] In summary, while existing flexible actuation technologies have made significant progress along multiple paths, none have broken through the technological paradigm of "single functional unit": The first type, based on disordered materials, exhibits isotropic deformation, requiring external constraints or complex control to transform into directional motion, lacking intrinsic deformation-guiding capabilities; the second type, based on intrinsically anisotropic materials, while achieving directional actuation, still relies on homogeneous material units, with their deformation patterns uniquely determined by the material's intrinsic properties, failing to achieve spontaneous differentiation of different types of actuation functions at the same structural level; the third type, based on external constraints, has motion patterns fixed by the external structure, lacking reconfigurability, and similarly remains at the level of passively guiding a single material unit. All three types of technologies fail to achieve functional differentiation and spatial synergy of actuation units at the intrinsic material level, meaning they cannot enable different units to possess different actuation response characteristics at the material level, nor can they achieve endogenous spatial synergy of different actuation forces through structural topological arrangement. Summary of the Invention

[0007] The purpose of this application is to provide a flexible biomimetic muscle complex and its control method, drive module and robot, which can at least overcome the defects of the existing flexible drive technology.

[0008] In a first aspect, this application provides a flexible biomimetic muscle composite, comprising:

[0009] The system comprises multiple strip-shaped components, each made of an electrically actuated material capable of active deformation under electric field excitation, and exhibiting anisotropic deformation behavior under electric field excitation. These multiple strip-shaped components are divided into multiple component groups, each component group including at least one strip-shaped component. The multiple component groups include at least a first component group and a second component group, with the axial stiffness of the first component group being higher than that of the second component group. The first component group and the second component group are coupled through a pre-defined force transmission structure, such that, in response to electric field excitation, the first driving force generated by the first component group, including a first deformation, and the second driving force generated by the second component group, including a second deformation, spatially coordinate to achieve motion output from the flexible biomimetic muscle complex that simulates the movement characteristics of biological muscles.

[0010] In one possible implementation, the difference in axial stiffness between the first component group and the second component group is achieved by the ratio of the elastic modulus of the first component group to the elastic modulus of the second component group being greater than 3:1 and less than 20:1.

[0011] In one possible implementation, the difference in axial stiffness between the first component group and the second component group is achieved by the different cross-sectional geometric dimensions of the strip-shaped components of the first component group and the second component group.

[0012] In one possible implementation, the different cross-sectional geometric dimensions of the strip members in the first component group and the second component group include: the ratio of the cross-sectional diameters of the strip members in the first component group to those in the second component group is greater than 1.2:1 and less than 3:1.

[0013] In one possible implementation, the difference in axial stiffness between the first component group and the second component group is achieved by the different products of the elastic modulus and cross-sectional area of ​​the strip-shaped components of the first component group and the second component group.

[0014] In one possible implementation, the difference in axial stiffness between the first component group and the second component group is achieved by the difference in length of the strip-shaped components in the first component group and the second component group.

[0015] In one possible implementation, the difference in axial stiffness between the first component group and the second component group causes the first component group to undergo a first deformation under electric field excitation, the first deformation including at least contraction or elongation, and causes the second component group to undergo a first contraction or first elongation deformation along its axial direction, or a second contraction or second elongation deformation, under electric field excitation.

[0016] In one possible implementation, the second component group is configured to generate a first contraction deformation along the axial direction under electric field excitation, which, together with the first deformation, completes the motion output.

[0017] In one possible implementation, the second component group is configured to generate a second axial contraction deformation under electric field excitation, which spatially counteracts the first deformation of the first component group and cooperates to complete the motion output.

[0018] In one possible implementation, the first component group and the second component group are each controlled by independent electrode circuits. In response to electric field excitation with different start times and / or different durations, the contraction deformation of the first component group and the second component group forms a preset sequential or superimposed relationship in time, thereby achieving smooth establishment and attenuation of contraction force during the axial contraction process of the composite.

[0019] In one possible implementation, the first deformation includes at least a contraction or elongation motion, and the second deformation includes at least a bending motion.

[0020] In one possible implementation, the first deformation and the second deformation work together to produce torsional motion, spiral motion, oscillating motion, or wave motion.

[0021] In one possible implementation, the first component group and the second component group respond to electric field excitation with different start times, different durations or different frequencies, such that the first deformation and the second deformation form a preset phase relationship, sequence relationship or frequency combination in time, thereby enabling the flexible biomimetic muscle complex to generate the motion output.

[0022] In one possible implementation, the strip-shaped members constituting the first component group and / or the second component group are configured to be independently addressable and driven, and the motion output is achieved by applying differentiated driving electric fields to different strip-shaped members or different component groups.

[0023] In one possible implementation, the first component group and the second component group have a preset first spatial structural relationship, which includes at least one of the following: first attachment type: at least one strip-shaped component in the second component group is attached longitudinally to the surface of at least one strip-shaped component in the first component group; first covering type: at least one strip-shaped component in the second component group covers the outer periphery of at least one strip-shaped component in the first component group in a spiral or woven manner; first embedding type: at least one strip-shaped component in the second component group is embedded inside at least one strip-shaped component in the first component group to form a skin-core structure; first spacing type: the first component group and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction.

[0024] In one possible implementation, the strip-shaped members in the first component group and the strip-shaped members in the second component group are arranged in three-dimensional space along a preset spiral path. The strip-shaped members in the first component group are arranged along the main direction of the spiral to provide axial contraction constraint, and the strip-shaped members in the second component group are arranged in a cross spiral to guide torsional motion, so that the composite generates a torsional moment about its axis under electric field excitation.

[0025] In one possible implementation, at least a portion of the strip-shaped members in the second component group are made of a conductive polymer, and their driving voltage is lower than that of the first component group.

[0026] In one possible implementation, the first component group and the second component group are arranged in an antagonistic pair in space, such that when both are excited at the same time, the flexible biomimetic muscle complex produces a rigidity enhancement effect.

[0027] In one possible implementation, the plurality of component groups further includes a third component group, the elastic modulus of which is greater than that of the second component group and less than that of the first component group; the first component group, the third component group, and the second component group have a preset second spatial structural relationship, the second spatial structural relationship including at least one of the following: second attachment type: the third component group is attached to the first component group, and the second component group is attached to the third component group; second covering type: the third component group covers the first component group, and the second component group covers the third component group; second embedded type: the third component group is embedded inside the first component group, and the second component group is embedded inside the third component group; second spaced type: the first component group, the third component group, and the second component group are arranged alternately at intervals in a direction perpendicular to the longitudinal direction.

[0028] In one possible implementation, the elastic moduli of the first component group, the third component group, and the second component group are distributed in a continuous gradient along the radial direction on the cross-section of the composite.

[0029] In one possible implementation, a flexible strain sensor integrated thereon is also included for real-time monitoring of the deformation of the composite and transmitting feedback signals to the control system.

[0030] In one possible implementation, at least a portion of the multiple strip-shaped components is an integrated drive-sensing component, which specifically includes at least one of the following: a first drive-sensing component, composed of a piezoelectric material and an electro-actuating material, configured to acquire the piezoelectric signal generated by the deformation in real time while undergoing active deformation under electric field excitation; and a second drive-sensing component, composed of an ionic electro-actuating material, with an electrode array disposed inside or on its surface, configured to monitor the change in ion concentration distribution representing the deformation state of the component in real time when undergoing active deformation under electric field excitation.

[0031] In one possible implementation, the strip-shaped component itself has a multi-level structure, including: a first-level structure comprising nanoscale myofibril biomimetic units composed of oriented functional fillers or molecular chains; and a second-level structure comprising micron-level myofibril biomimetic units formed by bundles of multiple said myofibril biomimetic units, the micron-level myofibril biomimetic units constituting the basic units of the first component group or the second component group.

[0032] In one possible implementation, the strip-shaped member includes reinforcing structures disposed at both ends, the reinforcing structures comprising at least one of the following: a locally thickened structure, wherein the cross-sectional area of ​​the strip-shaped member at both ends is larger than that in the middle region; a gradient modulus structure, wherein the elastic modulus of the strip-shaped member at both ends is higher than that in the middle region, forming a mechanical gradient increasing from the middle to both ends; a fiber reinforcement layer, wherein a high-strength fiber reinforcement layer or a mesh constraint layer is wrapped around the outer periphery of the end regions of the strip-shaped member; and an anchoring interface reinforcement layer, wherein an interface reinforcement layer integrally formed with the force transmission structure is disposed at both ends of the strip-shaped member.

[0033] In one possible implementation, the preset biomimetic arrangement configuration includes at least one of a biomimetic weaving configuration, a biomimetic feather configuration, a radial arrangement configuration, or a ring arrangement configuration.

[0034] In one possible implementation, the preset force transmission structure includes a biomimetic arrangement configuration, which is a biomimetic feather configuration. The first component group is arranged along the functional axis of the composite, and the second component group is arranged at a preset angle to the functional axis. The range of the preset angle is consistent with the physiological range of the feather angle of the muscle fibers of the target pennate or semi-pennate muscle.

[0035] In one possible implementation, the first component group and / or the second component group are divided into multiple independently replaceable fiber modules. Each fiber module is composed of a bundle of strip-shaped components with a preset elastic modulus value, and the ends of the fiber modules are provided with mechanical interfaces and electrical connection interfaces. The fiber modules are detachably assembled through the mechanical interfaces, and the electrical connection interfaces are used to provide a driving electric field for the strip-shaped components within the fiber modules and to transmit sensing signals when the strip-shaped components have sensing functions. By selecting fiber modules with different modulus values ​​for combination, or changing the radial arrangement order of the fiber modules on the cross-section of the flexible biomimetic muscle composite, the modulus spatial distribution of the flexible biomimetic muscle composite can be adjusted, thereby reconstructing its macroscopic motion output characteristics.

[0036] In one possible implementation, the multiple strip-shaped components are arranged in space to form at least two component groups with different extension directions, and the component groups are force-coupled through shared anchor points, interlacing connection points or flexible bases.

[0037] In one possible implementation, the multiple strip-shaped members are arranged in space in at least one of the following forms: parallel arrangement; hexagonal close-packed arrangement; nested arrangement in which a large-diameter strip-shaped member is nested within one or more small-diameter strip-shaped members; at least a portion of the strip-shaped members remain in a bent state in the bundled state, with a bending angle of 0°-180°; at least a portion of the strip-shaped members remain in a torsional state in the bundled state, with a torsional angle of 0°-360°.

[0038] In one possible implementation, the anisotropy of the strip-shaped member is achieved through at least one of the following methods:

[0039] The polymer chains inside the strip-shaped component are oriented longitudinally.

[0040] The strip-shaped component contains directionally arranged functional fillers;

[0041] The strip-shaped component is prepared and formed using a unidirectional stretching process;

[0042] The strip-shaped component is made by cutting anisotropic material;

[0043] The strip-shaped component is provided with an external constraint structure, which includes a fiber reinforcement layer, a mesh constraint layer, or discrete constraint elements.

[0044] In one possible implementation, the ratio of the main deformation along the longitudinal extension direction to the deformation along the transverse direction of the strip member under electric field excitation is greater than 5:1.

[0045] In one possible implementation, the electro-actuating material is selected from at least one of dielectric elastomers, ionomer-metal composites, liquid crystal elastomers, conductive polymers, or composites thereof; or, the strip member is composed of a thermally actuated material, a magnetostrictive material, or a photoactuated material.

[0046] In one possible implementation, the force transmission structure includes at least one of the following: interlacing connection point, anchoring interface, flexible substrate, friction contact surface, covalent cross-linking interface, or physical winding structure.

[0047] In one possible implementation, the motion output includes at least one of longitudinal contraction, multi-directional bending, and torsion, or any combination thereof.

[0048] In one possible implementation, at least a portion of the multiple strip-shaped members are configured to be independently addressable and driven, and the motion output is achieved by applying differentiated driving electric fields to different strip-shaped members or groups of members.

[0049] In one possible implementation, a flexible isolation layer is also included between the multiple strip-shaped members and / or between the member groups. The flexible isolation layer is a porous or mesh structure, used to reduce frictional losses between adjacent strip-shaped members and / or to serve as a channel for containing and transporting functional fluids.

[0050] In one possible implementation, the composite is configured to achieve independent and continuous control of the composite motion output by adjusting the electrical signal parameters applied to the first component group and / or the second component group.

[0051] Secondly, this application also provides a flexible drive control method applied to the flexible biomimetic muscle complex provided in the first aspect. The method includes: a first component group responding to a first electric field excitation triggering the generation of a first driving force including a first deformation; a second component group responding to a second electric field excitation triggering the generation of a second driving force including a second deformation; based on the force transmission structure coupling between the first component group and the second component group, the first driving force and the second driving force cooperate with each other in space, so that the flexible biomimetic muscle complex simulates the motion output that conforms to the motion characteristics of biological muscles.

[0052] Thirdly, this application provides a method for preparing a flexible biomimetic muscle complex as described in any one of the first aspects, comprising the following steps:

[0053] Provide multiple strip-shaped members made of an electrically actuated material and having anisotropy, the strip-shaped members comprising at least a first group of members and a second group of members, wherein the elastic modulus of the first group of members is higher than that of the second group of members;

[0054] Based on the anatomical and functional parameters of the target organism's muscle, the arrangement parameters of the biomimetic arrangement configuration are determined. The functional parameters include the proportion and spatial distribution of different types of muscle fibers in the target muscle.

[0055] According to the arrangement parameters, the first component group and the second component group are coupled and shaped together through a force transmission structure to form a flexible biomimetic muscle complex.

[0056] In one possible implementation, the method employs multi-material 3D printing or 4D printing technology. Based on a preset digital model, at least two electro-actuated materials with different elastic moduli are integrally deposited and formed according to the preset biomimetic arrangement. During the printing process, anisotropic structures are induced to form inside the materials by controlling the printing path, curing conditions, or applying an external field.

[0057] Fourthly, this application provides a driver module, comprising:

[0058] Flexible substrate;

[0059] And at least one flexible biomimetic muscle complex as described in any of the first aspects, the complex being fixed to the flexible substrate.

[0060] Fifthly, this application provides a bionic robot, which integrates at least two drive modules as described in the fourth aspect.

[0061] By utilizing the pre-defined axial stiffness difference between the first and second component groups, the two groups of components exhibit different mechanical response characteristics at the intrinsic material level, and achieve spatial coupling through a force transmission structure. When an electric field excitation is applied, the two groups of components spontaneously generate first and second deformations based on their intrinsic characteristics, and naturally coordinate in space, thereby directly outputting a composite motion consistent with the characteristics of biological muscle movement. This solution breaks through the technical paradigm of "single functional unit," achieving functional differentiation and spatial coordination of the drive unit without relying on external constraints or complex control. It overcomes the shortcomings of the first type of disordered materials requiring external guidance, the second type of intrinsically heterogeneous materials having homogeneous units and relying on independent control, and the third type of externally constrained motion patterns being fixed. It realizes the simulation of the multi-type unit division of labor and cooperation motion mechanism in biological muscle from the intrinsic material level, significantly improving the integration, naturalness of response, and structural reliability of the drive system. Attached Figure Description

[0062] Figure 1 A simplified structural diagram of a flexible biomimetic muscle composite provided in one embodiment of this application;

[0063] Figure 2 This is a schematic diagram of a first attachment type structure provided in one embodiment of this application;

[0064] Figure 3 This is a schematic diagram of a first encapsulated structure provided in one embodiment of this application;

[0065] Figure 4 This is a schematic diagram of the structure of a first embedded system provided in one embodiment of this application;

[0066] Figure 5 This is a schematic diagram of a first spaced structure provided in one embodiment of the present application;

[0067] Figure 6 This is a schematic diagram of a spiral path arrangement structure provided in one embodiment of this application;

[0068] Figure 7 This is a schematic diagram of a multi-level strip structure provided in one embodiment of this application;

[0069] Figure 8 This is a schematic diagram illustrating the cross-sectional area difference setting in one embodiment of this application;

[0070] Figure 9 This is a schematic diagram of the fiber reinforcement layer configuration provided in one embodiment of this application;

[0071] Figure 10 This is a schematic diagram of the connection structure of a fiber module according to an embodiment of this application;

[0072] Figure 11 A schematic diagram of a nested arrangement provided in one embodiment of this application;

[0073] Figure 12 A schematic diagram of a pre-bending shape provided in one embodiment of this application;

[0074] Figure 13 A schematic diagram of a pre-torsion configuration provided in one embodiment of this application;

[0075] Figure 14 This is a schematic diagram of a flexible isolation layer provided in one embodiment of this application;

[0076] Figure 15 This is a schematic diagram of a second attachment type provided in one embodiment of this application;

[0077] Figure 16 This is a schematic diagram of a second encapsulated structure provided in one embodiment of this application;

[0078] Figure 17 This is a schematic diagram of a second embedded structure provided in one embodiment of this application;

[0079] Figure 18 This is a schematic diagram of a second spacer structure provided in one embodiment of this application;

[0080] Figure 19 This is a flowchart illustrating a flexible drive control method provided in one embodiment of this application. Detailed Implementation

[0081] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0082] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0083] It should be understood that the terms "upper", "lower", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0084] Figure 1 This is a simplified structural diagram of a flexible biomimetic muscle complex provided in one embodiment of this application.

[0085] Reference Figure 1 As shown, the flexible biomimetic muscle composite may include multiple strip-shaped members 100, which are made of an electrically actuated material that can undergo active deformation under electric field excitation, and the deformation behavior under electric field excitation is at least anisotropic.

[0086] In some embodiments, the multiple strip-shaped members 100 serve as the basic driving unit of a flexible biomimetic muscle composite, the core of which lies in being composed of an electro-actuating material and possessing anisotropic properties. The electro-actuating material is selected from at least one of dielectric elastomers, ionomer-metal composites, liquid crystal elastomers, or conductive polymers. These materials can generate reversible active deformation under electric field excitation through Maxwell stress, ion migration, or molecular conformational changes. Anisotropy is achieved through specific preparation processes, such as uniaxial stretching of initially isotropic material films followed by heat setting treatment (e.g., stretching ratio of 3-5 times, temperature 80-150℃) to orient the polymer chains along the stretching direction, or by inducing the directional alignment of functional fillers (e.g., carbon nanotubes, barium titanate nanowires) through magnetic / electric field induction, thereby enabling the material to exhibit differentiated Young's modulus and driving response in different directions.

[0087] In some embodiments, the deformation anisotropy of the strip member under other excitation methods may include anisotropy under thermal excitation, anisotropy under magnetostrictive excitation, and anisotropy under photoexcitation. The following provides a detailed description of the three excitation methods:

[0088] Anisotropy under thermal excitation: When strip-shaped components are made of liquid crystal elastomer / carbon nanotube composite materials, they can undergo thermally induced deformation through the Joule heating effect (electrothermal effect) or directly through electrothermal deformation induced by an electric field-induced molecular orientation. For example, when liquid crystal elastomer / carbon nanotube composite fibers that have been uniaxially stretched and oriented are heated by electricity, they undergo thermally induced shrinkage along the orientation direction (i.e., the axial direction of the strip-shaped component) under Joule heating; at the same time, under a high-voltage electric field, the material can also directly undergo electrothermal shrinkage through Maxwell stress. Both exhibit anisotropy dominated by the orientation direction.

[0089] Anisotropy under magnetostrictive excitation: When strip-shaped components are made of dielectric elastomers doped with magnetic nanoparticles (such as Fe3O4), they can undergo electrostrictive deformation under electric field excitation through Maxwell stress, and simultaneously undergo magnetostrictive deformation under an applied magnetic field due to the magnetostriction or magnetic force of the magnetic particles. For example, when Fe3O4 nanochains are oriented in an SBAS matrix, the resulting component undergoes electrostrictive deformation along the orientation direction under a voltage of 500V, and simultaneously undergoes magnetostrictive deformation along the same direction under an alternating magnetic field of 0.5T; both are anisotropic.

[0090] Anisotropy under photoexcitation: When the strip component is made of azobenzene-doped liquid crystal elastomer, it undergoes electro-induced deformation due to the reorientation of liquid crystal units under electric field excitation, and at the same time, it undergoes photo-induced deformation due to the photo-isomerization of azobenzene under linearly polarized light irradiation.

[0091] The following examples illustrate how "the anisotropy of strip-shaped components is achieved through specific manufacturing processes".

[0092] I. Room temperature stretching combined with heat relaxation process:

[0093] In one embodiment, taking SBAS triblock copolymer as an example, an initially isotropic film is uniaxially stretched to four times its length at room temperature, followed by thermal stress relaxation treatment at 120°C. This reconstructs the polystyrene phase region and fixes the molecular orientation, resulting in an anisotropic strip-shaped component with a modulus of approximately 0.95 MPa in the stretching direction and approximately 0.34 MPa in the perpendicular direction, with a modulus ratio of approximately 2.8:1. Under electric field excitation, the electrostrain in the orientation direction can reach more than five times that in the perpendicular direction. Through this intrinsic anisotropic design of the material, the strip-shaped component 100 can generate efficient directional deformation along a preset direction without external constraints, providing a controllable deformation basis for subsequent spatial coordination among multiple component groups. This significantly improves driving efficiency and motion accuracy, overcoming the shortcomings of traditional isotropic materials that require complex constraints or control to achieve directional driving.

[0094] II. Hot stretching combined with cold setting process:

[0095] In another embodiment, taking a polyurethane-based dielectric elastomer composite material as an example, thermoplastic polyurethane is first blended with oriented conductive fillers (such as silver-plated barium titanate nanowires), and an initially isotropic strip precursor is obtained by melt extrusion. The precursor is then heated above its glass transition temperature (approximately 80-100°C) to a highly elastic state, and uniaxially stretched to three times its length at this temperature, causing the polymer chains and fillers to align synergistically along the stretching direction. While maintaining the stretched state, the temperature is lowered to room temperature, effectively "freezing" and fixing the oriented structure. Testing shows that the strip component prepared by this process has a modulus of approximately 1.5 MPa in the orientation direction and approximately 0.5 MPa in the perpendicular direction, with a modulus ratio of 3:1. Under electric field excitation, the electrostrain in the orientation direction can reach more than four times that in the perpendicular direction. Through this "hot stretching and cold setting" preparation method, the strip component 100 can also obtain a stable anisotropic structure. It can generate efficient directional deformation along the preset direction without external constraints, providing a controllable deformation basis for the subsequent spatial coordination between multiple component groups. This significantly improves driving efficiency and motion accuracy, overcoming the defect that traditional isotropic materials need to rely on complex constraints or control to achieve directional driving.

[0096] III. Magnetic field-induced directional alignment process:

[0097] In another embodiment, taking a composite material of liquid crystal elastomer and magnetic nanoparticles (such as iron oxide nanoparticles) as an example, the magnetic nanoparticles are first uniformly dispersed in the liquid crystal elastomer prepolymer, injected into a mold, and then placed in a strong magnetic field environment (such as a uniform magnetic field of 1.5T) for cross-linking polymerization. Under the action of the magnetic field, the magnetic nanoparticles drive the liquid crystal units to align in the direction of the magnetic field, forming an anisotropic structure. After cross-linking and curing, the magnetic field is removed, and a strip-shaped component is obtained. Testing shows that the modulus of this component along the orientation direction is approximately 1.1 MPa, and the modulus in the perpendicular direction is approximately 0.35 MPa, with a modulus ratio of approximately 3.1:1. Under electric field excitation, the electrostriction in the orientation direction can reach more than 4.5 times that in the perpendicular direction. Through this "magnetic field-induced directional alignment" preparation method, the strip component 100 can also obtain a stable anisotropic structure. It can generate efficient directional deformation along a preset direction without external constraints, providing a controllable deformation basis for subsequent spatial coordination among multiple component groups. This significantly improves driving efficiency and motion accuracy, overcoming the shortcomings of traditional isotropic materials that require complex constraints or control to achieve directional driving.

[0098] 4. Electric field-induced directional alignment process:

[0099] In another embodiment, taking a composite material of a dielectric elastomer and a high-dielectric-constant filler (such as barium titanate nanowires) as an example, the barium titanate nanowires are first dispersed in a polydimethylsiloxane prepolymer, injected into a mold, and then placed in a high-voltage electrostatic field (such as an electric field of 10 kV / cm) for cross-linking and curing. Under the action of the electric field, the high-dielectric-constant filler forms a chain-like arrangement along the direction of the electric field, imparting anisotropy to the material. After cross-linking and curing, the electric field is removed, resulting in a strip-shaped component. Testing showed that the modulus of this component along the orientation direction is approximately 0.9 MPa, and the modulus in the perpendicular direction is approximately 0.3 MPa, with a modulus ratio of approximately 3:1. The electrostriction in the orientation direction under electric field excitation can reach more than four times that in the perpendicular direction. Through this "electric field-induced directional alignment" preparation method, the strip component 100 can also obtain a stable anisotropic structure, which can generate efficient directional deformation along the preset direction without external constraints. This provides a controllable deformation basis for the subsequent spatial coordination between multiple component groups, significantly improving driving efficiency and motion accuracy, and overcoming the defect that traditional isotropic materials need to rely on complex constraints or control to achieve directional driving.

[0100] 5. Shear flow field directional alignment process:

[0101] In another embodiment, taking a composite material of thermoplastic elastomer and fibrous filler (such as carbon nanotubes) as an example, carbon nanotubes are dispersed in molten thermoplastic elastomer and extruded through a slit extrusion die. During extrusion, the melt is subjected to a strong shear flow field within the die, causing the carbon nanotubes to align in the direction of flow. After extrusion and cooling, an anisotropic strip-shaped component is obtained. Testing shows that the component has a modulus of approximately 1.3 MPa along the extrusion direction and approximately 0.4 MPa in the vertical direction, with a modulus ratio of approximately 3.25:1. Under electric field excitation, the electrostrain in the orientation direction can reach more than five times that in the vertical direction. Through this "shear flow field oriented alignment" preparation method, the strip-shaped component 100 can also obtain a stable anisotropic structure, generating efficient directional deformation along a preset direction without external constraints. This provides a controllable deformation basis for subsequent spatial coordination among multiple component groups, significantly improving driving efficiency and motion accuracy, and overcoming the shortcomings of traditional isotropic materials that require complex constraints or control to achieve directional driving.

[0102] 6. Photo-alignment preparation process:

[0103] In another embodiment, taking a liquid crystal elastomer containing photosensitive groups as an example, after the material is injected into a mold, it is irradiated with linearly polarized ultraviolet light (wavelength 365nm, light intensity 50mW / cm²). Under the action of polarized light, the photosensitive groups undergo a photochemical orientation reaction, driving the liquid crystal units to align in the polarization direction; subsequently, the orientation structure is fixed by thermal crosslinking or photocrosslinking to obtain a strip-shaped component. Testing shows that the modulus of this component along the orientation direction is approximately 0.85MPa, and the modulus in the perpendicular direction is approximately 0.28MPa, with a modulus ratio of approximately 3.03:1. Under electric field excitation, the electrostrain in the orientation direction can reach more than 4.2 times that in the perpendicular direction. Through this "photo-oriented" preparation method, the strip-shaped component 100 can also obtain a stable anisotropic structure, generating efficient directional deformation along a preset direction without external constraints. This provides a controllable deformation basis for subsequent spatial coordination between multiple component groups, significantly improving driving efficiency and motion accuracy, and overcoming the shortcomings of traditional isotropic materials that require complex constraints or control to achieve directional driving.

[0104] It should be noted that the various preparation methods listed above, such as room temperature stretching combined with thermal relaxation, hot stretching combined with cold setting, magnetic field-induced directional alignment, electric field-induced directional alignment, shear flow field-induced directional alignment, and photo-oriented alignment, are merely illustrative examples of obtaining anisotropic properties in the strip-shaped components of this application, and are not intended to limit this application. By inducing anisotropic deformation behavior in the strip-shaped components under electric field excitation, those skilled in the art can, according to the actual material system, process conditions, and performance requirements, employ other equivalent preparation methods capable of achieving molecular chain orientation, directional arrangement of functional fillers, or the introduction of external constraint structures, such as mechanical scratch-induced orientation, template growth, and self-assembly technology. As long as the strip-shaped components can exhibit differentiated driving responses in different directions, they all fall within the protection scope of this application, and will not be elaborated upon here.

[0105] Reference Figure 1 As shown, in some embodiments, the multiple strip members 100 can be divided into multiple member groups, each member group including at least one strip member 100.

[0106] In some embodiments, the plurality of component groups include at least a first component group C1 and a second component group C2, wherein the axial stiffness of the first component group C1 is higher than that of the second component group C2.

[0107] Based on the aforementioned anisotropic strip-shaped components, this application further functionally divides them according to axial stiffness. It should be noted that this "division" does not alter the structure of the components themselves, but rather involves testing and grouping the fabricated strip-shaped components based on their axial stiffness: by testing the axial stiffness value of each component (determined by both the elastic modulus and cross-sectional dimensions), strip-shaped components with similar stiffness are grouped together, thus forming a set of strip-shaped components with different axial stiffness ranges within the flexible biomimetic muscle composite (hereinafter referred to as the composite). Each component group consists of at least one strip-shaped component, possessing a specific axial stiffness range and driving response characteristics; different component groups exhibit differentiated driving behavior tendencies through preset differences in axial stiffness. Component groups with higher axial stiffness naturally tend to produce dominant axial contraction under electric field excitation, while component groups with lower axial stiffness are more prone to bending or auxiliary deformation. Through this "functional differentiation" design based on axial stiffness grouping, the composite can achieve progressive functional enrichment from binary to multi-functional within the same structural framework. It directly simulates the natural division of labor mechanism between fast-twitch muscle fibers (high axial stiffness, strong contraction) and slow-twitch muscle fibers (low axial stiffness, fatigue resistance) in biological muscles. This makes the motion mode of the composite no longer dependent on the complexity of external control, but naturally generated through the intrinsic characteristics of the component group and subsequent spatial arrangement, which significantly improves the structural efficiency and biomimetic realism of the system.

[0108] When the application scenario places higher demands on motion complexity, a third component group, a fourth component group, etc., can be added as needed.

[0109] It should be noted that in the embodiments of this application, "axial stiffness" refers to the ability of a strip-shaped member or group of members to resist tensile or compressive deformation in the axial direction, and its value ( ) equals the elastic modulus ( ) and cross-sectional area ( The product of ) divided by the length ( ),Right now For a component group consisting of multiple strip-shaped components bundled together, its axial stiffness is the sum of the axial stiffnesses of each component.

[0110] In some embodiments, the difference in axial stiffness between the first component group and the second component group can be achieved in a variety of ways:

[0111] 1. Material path: Keep the cross-sectional dimensions of the two sets of components the same, and achieve the difference in axial stiffness by controlling the elastic modulus ratio to be greater than the elastic modulus ratio threshold.

[0112] 2. Geometric path:

[0113] ① Keep the materials of the two sets of components the same, and achieve the difference in stiffness by controlling the different geometric dimensions of the cross sections (such as the cross section diameter ratio of 1.2:1-3:1);

[0114] ② Keep the materials and cross-sectional geometric dimensions of the two sets of components the same, and achieve the difference in stiffness by controlling the length.

[0115] 3. Combination path: Simultaneously adjust the material modulus and cross-sectional dimensions to achieve axial stiffness differences by varying the product of the elastic modulus and the cross-sectional area.

[0116] The appropriate method can be selected according to actual needs. As long as the axial stiffness of the first component group is higher than that of the second component group, functional division based on axial stiffness can be achieved.

[0117] To verify the effect of axial stiffness differences on functional differentiation, three groups of samples were prepared: Group A (modulus ratio 3:1, same cross-section), Group B (same modulus, cross-section diameter ratio 2:1), and Group C (modulus ratio 2:1, cross-section diameter ratio 1.5:1, axial stiffness ratio 3:1). All three groups of samples exhibited clear functional differentiation under electric field excitation: the first group dominated axial contraction, while the second group dominated bending deformation, demonstrating that axial stiffness differences are the essential factor for functional differentiation.

[0118] In some embodiments, the difference in axial stiffness between the first component group and the second component group is achieved by the ratio of the elastic modulus of the first component group to the elastic modulus of the second component group being greater than 3:1 and less than 20:1.

[0119] In one embodiment, the difference in axial stiffness between the first and second component groups is achieved by a ratio of elastic modulus greater than 3:1 and less than 20:1. This range is an optimized parameter determined based on a comprehensive consideration of achieving stable functional differentiation and structural reliability. It should be noted that axial stiffness is determined by both the elastic modulus and the cross-sectional geometry. When the cross-sectional dimensions of the two component groups are the same, the difference in axial stiffness directly reflects the difference in elastic modulus. Extensive experimental research has shown that when the elastic modulus ratio is less than 3:1, the driving behaviors of the two component groups under electric field excitation tend to converge, failing to form a clear functional differentiation. For example, the first component group cannot dominate axial contraction, and the second component group cannot dominate bending deformation; both exhibit mixed deformation modes, making spatial coordination difficult. However, when the modulus ratio reaches 3:1, the axial strain of the first component group can reach more than five times its transverse strain, and the bending deflection of the second component group can reach more than three times its axial strain, forming a clear dominant driving functional differentiation. Taking SBAS material as an example, the modulus ratio can be precisely controlled by adjusting the stretching ratio: the first component group is stretched at a high ratio (stretching ratio 5:1) to obtain a modulus of 1.2 MPa, while the second component group is stretched at a low ratio (stretching ratio 2:1) to obtain a modulus of 0.4 MPa. The modulus ratio between the two is exactly 3:1, which meets the functional differentiation requirement. On the other hand, when the modulus ratio exceeds 20:1, the second component group is prone to buckling or excessive creep under electric field excitation due to its low stiffness, making it difficult to generate controllable bending driving force. At the same time, the excessive modulus difference will cause significant stress concentration at the interface, leading to premature interface failure of the composite during dynamic driving. Therefore, controlling the elastic modulus ratio between 3:1 and 20:1 can ensure that the first and second component groups form a sufficient axial stiffness difference under the premise of the same cross-sectional size. This ensures both the intrinsic functional differentiation of the material and the structural stability and interface reliability of the second component group, enabling the composite to maintain stable composite motion output during long-term dynamic driving.

[0120] It should be noted that, to clarify the impact of the difference in axial stiffness between the first and second component groups on the differentiation of driving behavior, this application established the critical modulus ratio for achieving functional differentiation of "contraction-dominant and bending-dominant" through a series of comparative experiments. Specifically, five composite sample groups with elastic modulus ratios (Ehigh / Elow) of 2.5:1, 2.8:1, 3.0:1, 3.2:1, and 3.5:1 were prepared. Except for the modulus ratio, the material composition, cross-sectional geometry, anisotropy, arrangement, and force transmission structure of the strip components in each sample group were kept consistent. Under the same electric field strength (e.g., 40V / μm), the axial strain εaxial of the first component group (high modulus group) and the maximum bending deflection ωbend of the second component group (low modulus group) were simultaneously acquired using digital image correlation method, and the functional differentiation coefficient η=εaxial / (ωbend / L) (where L is the component gauge length) was used as a quantitative evaluation index. Experimental results show that when the modulus ratio is ≤2.8:1, the deformation modes of the first and second component groups are similar, the ratio of the axial strain of the first component group to that of the second component group is less than 1.2, and the second component group does not show significant directional bending (η<0.5), indicating that both groups of components exhibit mixed deformation under the electric field and cannot form a clear functional division. When the modulus ratio reaches 3.0:1, the axial strain of the first component group increases sharply to more than 5 times its transverse strain, while the bending deflection of the second component group simultaneously reaches more than 3 times its axial strain, and the functional differentiation coefficient η jumps to more than 2.8. When the modulus ratio continues to increase to 3.2:1 and above, the differentiation effect is further enhanced and tends to stabilize. The experimental data fully demonstrates that 3:1 is the critical technical threshold for achieving the core functional differentiation of this application (i.e., the high-modulus group dominates axial contraction, and the low-modulus group dominates bending drive). When the modulus ratio is lower than this value, the mechanical responses of the two groups of components under electric field excitation tend to converge, making it impossible to achieve the cooperative motion basis required by this application. Only when the modulus ratio reaches or exceeds 3:1 can the intrinsic stiffness difference of the materials overcome the constraint coupling effect brought about by structural coupling, thereby triggering a clear functional division of "contraction and bending". In this embodiment, by controlling the elastic modulus ratio between 3:1 and 20:1, sufficient axial stiffness difference between the first component group and the second component group is achieved, providing a reliable mechanical basis for subsequent spatial coordination.

[0121] In some embodiments, the difference in axial stiffness between the first component group and the second component group is achieved by the difference in the cross-sectional geometry of the strip members of the first component group and the second component group. Specifically, the difference in the cross-sectional geometry of the strip members of the first component group and the second component group includes: the ratio of the cross-sectional diameter of the strip members of the first component group to that of the strip members of the second component group is greater than 1.2:1 and less than 3:1.

[0122] In one implementation, the difference in axial stiffness between the first and second component groups is achieved through the different cross-sectional geometries of their strip-shaped components. When both groups of components use the same material (i.e., the same elastic modulus), the axial stiffness is proportional to the cross-sectional area. Therefore, the required stiffness difference can be achieved independently by adjusting the cross-sectional dimensions without changing the material composition. This design approach is particularly suitable for scenarios with high requirements for material homogeneity or where a simplified material system is desired. For example, in medical applications with strict biocompatibility requirements, the same biosafety certified material can be used, and functional differentiation needs can be met solely through geometric design.

[0123] Under the premise of identical materials (i.e., equal elastic modulus), the difference in axial stiffness between the first and second component groups can be achieved by adjusting the cross-sectional geometry of the strip component. When the materials of the two groups of components ( ) and length ( When all cross-sectional areas are the same, the difference in axial stiffness depends entirely on the cross-sectional area. The difference. For example, for a circular cross-section member, the cross-sectional area... Proportional to the square of the diameter, a small difference in cross-sectional diameter is magnified to a squared difference in stiffness. Experiments show that when the ratio of the cross-sectional diameters of the first component group to the second component group is greater than 1.2:1 and less than 3:1, sufficient difference in axial stiffness can be achieved to realize a clear functional differentiation. Taking SBAS material as an example, if both groups of components use a material with a modulus of 0.4 MPa, the diameter of the first component group is 0.3 mm (cross-sectional area of ​​approximately 0.07 mm²), and the diameter of the second component group is 0.15 mm (cross-sectional area of ​​approximately 0.018 mm²). The diameter ratio of 2:1 corresponds to a cross-sectional area ratio of 4:1, and the axial stiffness ratio is also 4:1. Under these conditions, when excited by an electric field, the first component group, due to its higher axial stiffness, dominates axial contraction, while the second component group, due to its lower axial stiffness, is more prone to bending deformation. This difference is sufficient to overcome the constraint coupling effect caused by structural coupling, achieving a stable functional division of "contraction-dominant and bending-dominant". The optimal range for this diameter ratio is based on the following: when it is below 1.2:1, the stiffness difference is too small, and the driving behavior of the two groups of components tends to be similar; when it is above 3:1, the second group of components is prone to instability or fracture due to being too thin, and the stress concentration at the interface increases significantly. Therefore, controlling the cross-sectional diameter ratio between 1.2:1 and 3:1 can ensure functional differentiation while taking into account structural reliability and process feasibility.

[0124] In one implementation, the difference in cross-sectional geometry is specifically manifested in a cross-sectional diameter ratio greater than 1.2:1 and less than 3:1. Related experiments show that when the cross-sectional diameter ratio is less than 1.2:1, the difference in axial stiffness between the two groups of components is too small, and their driving behavior under electric field excitation tends to be similar: the first component group cannot dominate axial contraction, and the second component group cannot produce significant directional bending; both exhibit a mixed deformation mode, making it difficult to achieve the functional division of "contraction and bending" required for spatial coordination. However, when the cross-sectional diameter ratio reaches 1.2:1, since axial stiffness is proportional to the square of the diameter, the axial stiffness ratio of the two groups of components is approximately 1.44:1, which is sufficient to trigger a clear functional differentiation. Specifically, the axial strain of the first component group can reach more than four times its transverse strain, and the bending deflection of the second component group begins to appear and increases rapidly with the increase of the diameter ratio.

[0125] In some embodiments, the difference in axial stiffness between the first component group and the second component group is achieved by the difference in length of the strip members in the first component group and the second component group.

[0126] In a specific implementation, when the materials of the two sets of components ( ) and cross-sectional dimensions ( When all elements are the same length, axial stiffness is inversely proportional to length; that is, the shorter the component, the higher the axial stiffness, and the longer the component, the lower the axial stiffness. Therefore, by using shorter strip-shaped components in the first component group and longer strip-shaped components in the second component group, differences in axial stiffness can be achieved without changing the material and cross-section. For example, one way to achieve this is through differentiated anchor point design: the two ends of the strip-shaped component in the first component group are fixed to two anchor points that are close together (e.g., 50mm in length), and the two ends of the strip-shaped component in the second component group are fixed to two anchor points that are far apart (e.g., 80mm in length). Since both components have the same material and cross-section, the axial stiffness of the first component group is approximately 1.6 times that of the second component group (80 / 50). The second implementation method is segmented clustering within the component group: strip-shaped components of the same material are prepared with the same diameter, but the first component group is composed of 20 short components with a length of 60mm bundled in parallel, and the second component group is composed of 20 long components with a length of 100mm bundled in parallel. The stiffness ratio of the two groups of components is the inverse length ratio 100:60≈1.67:1.

[0127] In the flexion actuation application of bionic fingers, the actuating component of the proximal phalanx can be designed to be shorter to obtain high stiffness for strong flexion, while the actuating component of the distal phalanx can be designed to be longer to obtain low stiffness for fine bending. Although both are made of the same material, functional differentiation is achieved through the difference in length. This approach does not require changes to the material formula or cross-sectional dimensions; stiffness can be flexibly adjusted solely through length design. It is particularly suitable for applications where space is limited but the length direction can be freely adjusted.

[0128] In some embodiments, the difference in axial stiffness between the first component group and the second component group is achieved by the different products of the elastic modulus and cross-sectional area of ​​the strip members of the first component group and the second component group.

[0129] In one implementation, when the lengths of the two sets of components... When they are the same, the difference in their axial stiffness is directly determined by the elastic modulus. With cross-sectional area The product of the two components determines the axial stiffness. Therefore, the difference in axial stiffness between the first and second component groups can be adjusted. The product is achieved by combining the effects of both material and geometric dimensions. Specific implementation methods include, but are not limited to, the following three scenarios: First, keeping the cross-sectional areas of the two sets of components the same, and using materials with different moduli to achieve… The products are different (e.g., the first component group uses a high-modulus material, and the second component group uses a low-modulus material); secondly, keeping the materials of the two groups of components the same, and designing different cross-sectional dimensions to... Differences thus lead to The products are different (e.g., the first component group uses a large-diameter component, and the second component group uses a small-diameter component); thirdly, the material modulus and cross-sectional dimensions are adjusted simultaneously to make the combined product... The product shows differences. Taking SBAS material as an example, the first component group can use components with a modulus of 1.2 MPa and a diameter of 0.3 mm ( The second component group can use components with a modulus of 0.4 MPa and a diameter of 0.2 mm. A product ratio of approximately 6.8:1 between the two components results in a significant difference in axial stiffness. Regardless of the method used to achieve this, as long as the first component group... If the product is greater than that of the second component group, it can ensure the functional differentiation of the first component group dominating axial contraction and the second component group dominating bending deformation under electric field excitation.

[0130] It should be noted that the difference between the product of elastic modulus and cross-sectional area is essentially a mathematical definition of the difference in axial stiffness. According to mechanics of materials, axial stiffness... In the length of the two sets of components Under the same conditions, the difference in stiffness is directly caused by The product determines the outcome. Therefore, as long as the two sets of components... Different values ​​indicate differences in axial stiffness, sufficient to achieve functional differentiation. Those skilled in the art can adjust the material modulus, cross-sectional dimensions, or a combination of both according to actual needs to make the first component group... The product is greater than that of the second component group, thus achieving a higher axial stiffness for the first component group than for the second component group.

[0131] The following examples illustrate the functional classification of strip-shaped components based on their elastic modulus.

[0132] I. Classification method based on absolute modulus threshold:

[0133] By controlling the manufacturing process (such as uniaxial stretching ratio, heat setting temperature, and functional filler orientation), strip-shaped components with modulus values ​​distributed in different ranges can be obtained. Based on the driving force requirements of the target application, a corresponding modulus threshold range is set. For example, in applications requiring greater shrinkage force, strip-shaped components with modulus values ​​in the 1.0-1.5 MPa range are classified as the first component group C1, making them naturally inclined to axial shrinkage under electric field excitation; strip-shaped components with modulus values ​​in the 0.1-0.3 MPa range are classified as the second component group C2, making them more prone to bending deformation. Taking SBAS material as an example, the first component group C1 uses strip-shaped components that have undergone ultra-high stretching (stretching ratio 6:1) and high-temperature flexible setting treatment, achieving a modulus of 1.5 MPa; the second component group C2 uses strip-shaped components with low stretching (stretching ratio 1.5:1), achieving a modulus of approximately 0.2 MPa, with a modulus ratio of 7.5:1. Based on this, multiple component groups, including at least a first component group and a second component group, constitute the most basic functional binary structure, achieving an endogenous coordination of the two driving forces: contraction and bending. By setting a clear modulus threshold, it is ensured that the two component groups have significant differences in modulus, thus enabling functional differentiation under electric field excitation: the first component group naturally dominates axial contraction, and the second component group naturally dominates bending deformation. This division method is simple and direct, and easy to implement for industrial-scale batch screening and quality control.

[0134] In other applications requiring a balance between contractile force and flexural flexibility, a modulus of 0.8-1.2 MPa can be selected as the first component group and 0.3-0.6 MPa as the second component group, with the modulus ratio controlled between 2 and 4. By setting a clear modulus threshold, it is ensured that the two groups of components have modulus differences that match the application requirements, thereby achieving the preset functional differentiation under electric field excitation.

[0135] II. Classification based on preparation process parameters:

[0136] By controlling the manufacturing process (such as uniaxial tensile ratio, heat setting temperature, and orientation of functional fillers), strip-shaped components with modulus values ​​distributed in different ranges can be obtained. The modulus ranges of the two sets of components can be flexibly adjusted according to different combinations of process parameters.

[0137] In one embodiment, when using a uniaxial stretching process, strip members with a stretching ratio of 5:1 are classified into the first component group C1, so that their modulus value is in the range of 1.2-1.5MPa, and they naturally tend to produce axial shrinkage under electric field excitation; strip members with a stretching ratio of 2:1 are classified into the second component group C2, so that their modulus value is in the range of 0.3-0.5MPa, and they are more prone to bending deformation.

[0138] In another implementation, if a magnetic field-induced orientation process is used, the two groups of components can be distinguished by adjusting the magnetic field strength: the strip-shaped components prepared under a 1.8T magnetic field are classified as the first group, with a modulus value of 1.3MPa; the strip-shaped components prepared under a 0.6T magnetic field are classified as the second group, with a modulus value of 0.4MPa. Taking SBAS material as an example, the first group C1 uses strip-shaped components with a stretch ratio of 5:1, with a modulus value of 1.2MPa, while the second group C2 uses strip-shaped components with a stretch ratio of 2:1, with a modulus value of 0.4MPa, resulting in a modulus ratio of 3:1. If a combination of stretch ratios of 6:1 and 1.5:1 is used, the modulus ratio can be further increased to over 6:1. Directly linking the preparation process parameters with the modulus grouping eliminates the need for subsequent individual testing and screening, significantly improving production efficiency. Simultaneously, by flexibly selecting combinations of process parameters, precise control over the modulus difference between the two groups of components can be achieved, meeting the differentiated requirements for shrinkage force and bending flexibility in different application scenarios.

[0139] III. Division based on target application functional requirements:

[0140] By controlling the manufacturing process (such as uniaxial stretching ratio, heat setting temperature, and orientation of functional fillers), strip-shaped components with modulus values ​​distributed in different ranges can be obtained. Based on the desired motion mode of the composite, the required shrinkage force for the first component group C1 and the required flexural flexibility for the second component group C2 are pre-set, and the required modulus range is deduced through the material constitutive relationship. Examples from different application scenarios are provided below.

[0141] In the main drive joint application of the robot bionic hand, a large contractile force is required to realize the grasping function of the hand. A strip-shaped component with a modulus in the range of 1.2-1.8MPa can be selected as the first component group C1. At the same time, a moderate bending flexibility is required to realize finger bending. A strip-shaped component with a modulus in the range of 0.2-0.4MPa can be selected as the second component group C2.

[0142] In facial expression robot applications, precise contraction control and small-amplitude bending are required. Therefore, a modulus of 0.5-0.8 MPa can be selected as the first component group, and 0.1-0.2 MPa as the second component group. Taking SBAS material as an example, for high-output scenarios, the first component group C1 uses strip-shaped components that have undergone high-ratio stretching (stretch ratio 5:1) and added rigid fillers, achieving a modulus of up to 1.5 MPa; the second component group C2 uses pure material components that have undergone low-ratio stretching (stretch ratio 2:1), with a modulus of approximately 0.3 MPa and a modulus ratio of 5:1.

[0143] For high-precision applications, the first component group C1 uses strip-shaped components with a tensile ratio of 3:1 and a modulus of 0.7 MPa; the second component group C2 uses strip-shaped components with a tensile ratio of 1.5:1 and a modulus of 0.15 MPa, with a modulus ratio of approximately 4.7:1.

[0144] By adopting the above-mentioned "as-needed division" method, the functional differentiation of the composite is highly matched with the target application, maximizing the utilization of material properties and achieving precise customization of driving performance.

[0145] IV. Division method based on modulus ratio:

[0146] By controlling the manufacturing process (such as uniaxial tensile ratio, heat setting temperature, and functional filler orientation), strip-shaped components with modulus values ​​distributed in different ranges can be obtained. Using a certain benchmark modulus value as a reference, a target ratio is set to divide the components into two groups. In one embodiment, multiple SBAS strip-shaped components from a batch are sampled and tested, and the average modulus is measured to be 0.5 MPa, which is then determined as the benchmark modulus for the current multiple strip-shaped components. That is, the reference modulus is 0.5 MPa.

[0147] In scenarios requiring strong contraction and weak bending, strip-shaped components with modulus values ​​more than three times the reference value can be classified into the first component group C1, and strip-shaped components with modulus values ​​less than one-third of the reference value can be classified into the second component group C2. Taking a reference modulus of 0.5 MPa as an example, when using three times the reference modulus and one-third of the reference modulus as the dividing thresholds, the modulus of the first component group C1 is ≥1.5 MPa, and the modulus of the second component group C2 is ≤0.17 MPa, thus achieving a modulus ratio of ≥9:1 between the first and second component groups.

[0148] In scenarios requiring balanced output, 2 times the baseline modulus and 1 / 2 the baseline modulus can be selected as the division thresholds. Taking a baseline modulus of 0.5 MPa as an example, when using 2 times the baseline modulus and 1 / 2 the baseline modulus for division, the modulus of the first component group is ≥1.0 MPa, and the modulus of the second component group is ≤0.25 MPa, thus achieving a modulus ratio of ≥4:1 between the first and second component groups. Taking SBAS material as an example, for high-proportion requirements, the first component group C1 uses strip-shaped components with a tensile ratio of 6:1 and added rigid fillers, achieving a modulus of up to 1.8 MPa; the second component group C2 uses foamed lightweight components with a tensile ratio of 1.2:1, achieving a modulus of approximately 0.1 MPa, resulting in a modulus ratio of 18:1. Dividing by proportional relationships can adapt to material fluctuations in different batches and different material systems, ensuring that the two component groups always have a preset modulus difference multiple, thereby stably achieving the required functional differentiation in different application scenarios and enhancing the versatility and robustness of the technical solution.

[0149] It should be noted that the modulus values, tensile ratios, and modulus difference factors of the first and second component groups in the above examples are provided as examples for different application scenarios and are not intended to limit the present invention. This application can also adaptively adjust and optimize the modulus range, process parameters, and proportional relationships according to actual needs, such as target shrinkage force, bending flexibility requirements, response speed, and energy consumption limitations. As long as the first and second component groups can form a preset difference in axial stiffness to achieve functional differentiation, they fall within the protection scope of this invention.

[0150] In some embodiments, the first component group and the second component group are coupled through a preset force transmission structure, such that when in response to electric field excitation, the first driving force generated by the first component group including the first deformation and the second driving force generated by the second component group including the second deformation cooperate with each other in space, so as to realize the motion output of the flexible biomimetic muscle complex that matches the motion characteristics of biological muscles.

[0151] In one embodiment, the first component group C1 and the second component group C2 are coupled through a preset force transmission structure, thereby achieving spatial coordination between the corresponding first and second deformations. When an electric field excitation is applied, the first component group C1, based on its high modulus characteristics, undergoes axial contraction deformation, generating a first driving force; the second component group C2, based on its low modulus characteristics, undergoes bending deformation, generating a second driving force. Since the two are tightly coupled through the force transmission structure, the axial contraction of the first driving force applies axial constraint to the second component group, guiding its bending deformation towards the target motion; simultaneously, the bending effect of the second driving force causes the contraction path of the first component group to deflect spatially. The two naturally form a composite motion through mutual constraint and cooperation. For example, in the application of robotic fingers, this coordination allows the composite to simultaneously achieve axial shortening of the phalanx and bending rotation around the joint to simulate the flexion action of a real finger. This spatial coordination mechanism based on structural coupling enables the composite to spontaneously generate composite motion at the moment of excitation without external control commands. It realizes a direct mapping from the intrinsic properties of materials to macroscopic motion output, significantly improving the integration and naturalness of the drive system, and overcoming the shortcomings of traditional technologies that rely on complex control algorithms or external constraints to achieve composite motion.

[0152] From a biomimetic perspective, even seemingly "simple" movements in biological muscles (such as maintaining posture or uniform contraction) are achieved through a synergistic mechanism where slow-twitch muscle fibers provide a stable foundation, while fast-twitch muscle fibers provide dynamic regulation. This synergistic architecture of multiple unit types endows biological muscles with superior stability, gradual flexibility, and precise control capabilities, not just the ability to generate complex movements. Existing technologies, lacking simulation of this underlying synergistic architecture, result in a fundamental gap between flexible actuation systems and biological muscles in terms of stability, disturbance resistance, and control precision. The flexible biomimetic muscle complex provided in this application achieves complex motion output through a multi-component design and a spatial synergistic mechanism.

[0153] The force transmission structure between the first component group C1 and the second component group C2 can take many forms. The following describes in detail the various forms of force transmission structures.

[0154] In some embodiments, the first component group and the second component group have a preset first spatial structure relationship, which includes at least one of the following: first attachment type, first covering type, first embedded type, and first spaced type.

[0155] First attachment type: At least one strip-shaped member in the second component group is attached longitudinally to the surface of at least one strip-shaped member in the first component group.

[0156] Figure 2 This is a schematic diagram of a first attachment structure provided in one embodiment of this application.

[0157] In one implementation, reference Figure 2 As shown, the first attachment method refers to at least one strip-shaped member in the second component group C2 being attached longitudinally (i.e., axially) to the surface of at least one strip-shaped member in the first component group C1, with force transmission achieved through interfacial bonding. In a specific implementation, an uncured slurry 201 of the same material or an interfacial coupling agent can be coated on the surface of the strip-shaped member of the first component group, and then the strip-shaped member of the second component group can be attached longitudinally thereon, forming an integrated interface through co-curing or chemical cross-linking.

[0158] For example, taking SBAS material as an example, a high-modulus (1.2 MPa) first component group strip is used as a substrate, and a low-modulus (0.4 MPa) SBAS prepolymer slurry is coated on its surface. Then, another low-modulus strip is attached to it, and co-cured at 120°C for 30 minutes, allowing the molecular chains at the interface to diffuse and entangle with each other, forming a strong chemical bond. When an electric field is applied, the first component group generates a contraction driving force along the axial direction, and the second component group generates a bending driving force. Since the two are tightly coupled through the attachment interface, the bending deformation of the second component group will directly act on the first component group, causing its contraction path to deflect. Meanwhile, the contraction constraint of the first component group will restrict the free bending of the second component group. The two naturally form a composite motion through mutual restraint.

[0159] In robotic bionic finger applications, this attachment structure allows contractile and bending forces to be transmitted synergistically along the longitudinal direction of the components, enabling flexion movements of the finger joints. This first type of attachment spatial structure achieves efficient force transmission between the two sets of components through simple surface attachment, eliminating the need for complex weaving or embedding processes, and making the fabrication process simple and controllable. Furthermore, the attachment interface can be designed as continuous contact or discrete point contact as needed, thereby adjusting the coupling strength between the two sets of components and achieving precise control over composite motion modes.

[0160] First covering type: At least one strip-shaped member in the second component group covers the outer periphery of at least one strip-shaped member in the first component group in a spiral or woven manner.

[0161] Figure 3 This is a schematic diagram of a first encapsulated structure provided in one embodiment of this application.

[0162] In one implementation, reference Figure 3As shown, the first type of covering refers to at least one strip-shaped member in the second component group C2 covering the outer periphery of at least one strip-shaped member in the first component group C1 in a spiral or woven manner, with force transmission achieved through interface contact and geometric constraints. In specific implementations, the high-modulus strip-shaped member of the first component group can be used as the core layer, and the low-modulus strip-shaped member of the second component group can be spirally wound around its outer surface with a certain pitch and covering angle. The number of winding layers can be designed as a single layer or multiple layers as needed; alternatively, a weaving process can be used to cover the outer periphery of the first component group with multiple strip-shaped members of the second component group in a warp and weft interlacing manner to form a mesh covering layer.

[0163] In some implementations, refer to Figure 3 As shown, to enhance interfacial bonding, coupling agent 301 or a slurry of the same material can be coated on the surface of the first component group before coating, and co-curing treatment can be performed after coating.

[0164] For example, taking SBAS material as an example, a first component group of strip-shaped members with a diameter of 0.5 mm and a modulus of 1.2 MPa is used as the core layer. A second component group of strip-shaped members with a diameter of 0.2 mm and a modulus of 0.4 MPa is spirally wound onto its surface with a 45° wrap angle and a pitch of 2 mm. After two layers are wound, a dilute SBAS solution is applied, and the mixture is dried and cured at 100°C, so that the interface between the wound layer and the core layer forms a chemical bond. When an electric field is applied, the first component group generates a contraction driving force along the axial direction, and the second component group generates a contraction driving force along the helical direction due to its spiral wrapping shape. This driving force can be decomposed into an axial component and a circumferential component. The circumferential component applies a torsional torque to the core layer, causing the axial contraction of the core layer to be accompanied by a torsional motion around the axis. The two work together to generate a helical motion.

[0165] In the application of bionic wrist joints in robots, this first encapsulated composite structure enables the composite to rotate while contracting axially, simulating the combined flexion and rotation of the wrist. Through helical or woven encapsulation, this composite structure transforms the bending drive of the second component group into a multidimensional force acting on the first component group, naturally coupling the originally singular contractile and torsional motions to directly output a composite motion. Simultaneously, by adjusting the encapsulation angle, pitch, and number of winding layers, the proportional relationship between axial contraction and torsion can be precisely controlled, enabling flexible design of motion patterns. Furthermore, the encapsulation structure creates a multi-layered composite in the radial direction, enhancing the overall structural integrity and fatigue resistance of the composite.

[0166] First Embedding: At least one strip-shaped component in the second component group is embedded inside at least one strip-shaped component in the first component group to form a core-skin structure.

[0167] Figure 4 This is a schematic diagram of a first embedded structure provided in one embodiment of this application.

[0168] In one implementation, reference Figure 4 As shown, the first embedded structure refers to at least one strip-shaped component in the second component group C2 being embedded inside at least one strip-shaped component in the first component group C1, forming a core-skin structure. In a specific implementation, a co-extrusion process can be used to simultaneously form two materials with different moduli. The high-modulus first component group C1 material is used as the core layer, and the low-modulus second component group C2 material is used as the skin layer. The skin layer material is used to cover and partially embed into the surface of the core layer through a co-extrusion die, forming a semi-embedded structure with a continuous core layer and an embedded skin layer. Alternatively, a casting process can be used. First, a strip-shaped component of the first component group with longitudinal grooves or channels is prepared as a substrate. Then, the low-modulus second component group material is injected into the grooves or channels in a molten state. After cooling, the two form a mechanical interlock and interface bond.

[0169] In one embodiment, the specific implementation of simultaneously molding two materials with different moduli using a co-extrusion process includes: using a high-modulus SBAS material with a modulus of 1.2 MPa as the core layer material and a low-modulus SBAS material with a modulus of 0.4 MPa as the skin layer material, both are fed to a co-extrusion die through two extruders; the die is designed with three circumferentially evenly distributed flow channels, allowing the skin layer material to be embedded in the surface of the core layer as three longitudinal ribs, forming a composite strip-shaped component with a continuous core layer and three low-modulus ribs embedded in the surface after extrusion. The extrusion temperature is controlled at 150°C, and the extrusion speed ratio is core layer:skin layer = 3:1. After cooling, an integrated composite component with a diameter of 1.0 mm and a rib depth of 0.2 mm is obtained. In this composite component, the high-modulus core layer constitutes the first component group, and the low-modulus ribs constitute the second component group. The two form an integrated interface during the molding process, requiring no subsequent bonding treatment. When an electric field is applied, the high-modulus core layer dominates axial contraction, while the embedded low-modulus ribs dominate bending deformation; the two work together naturally in three-dimensional space. Through co-extrusion, precise positioning and integrated molding of the two components at the microscale can be achieved, significantly improving interfacial bonding strength and force transmission efficiency, while simplifying the manufacturing process and making it suitable for mass production.

[0170] For example, taking SBAS material as an example, a first component group of strip-shaped components with a diameter of 0.8 mm and a modulus of 1.2 MPa is first prepared as a core layer by co-extrusion molding. Three V-shaped grooves with a depth of 0.2 mm are machined on its surface along the axial direction. Then, a second component group of material with a modulus of 0.4 MPa is heated to a molten state, injected into the grooves, and pressure is applied to make it fit tightly against the inner wall of the grooves. After cooling, the second component group of material solidifies to form three strip-shaped structures embedded in the core layer. When an electric field is applied, the first component group, as a continuous matrix, generates a contraction driving force along the axial direction, while the second component group, as an insert, generates a bending driving force due to its low modulus characteristics. Since the second component group is embedded inside the first component group, its bending deformation is constrained by the surrounding matrix in all directions and cannot bend freely. Instead, it applies local stress to the matrix, inducing the matrix to produce a composite bending in three-dimensional space.

[0171] In bionic finger applications, the composite structure constructed through this first embedded component enables the finger to achieve fine bending in multiple directions while contracting axially, simulating the combined flexion and lateral swing motion of a real finger. This composite structure uses high-modulus (1.2 MPa) SBAS material as the first component group, forming the core skeleton of the finger; low-modulus (0.4 MPa) SBAS material is used as the second component group, uniformly embedded circumferentially on the core surface in the form of three longitudinal ribs, corresponding to the palmar, dorsal, and lateral positions of the finger, respectively. When an electric field is applied, the high-modulus core layer dominates axial contraction, generating the main driving force for finger flexion; the low-modulus ribs embedded in different circumferential positions exhibit differentiated bending deformation due to their low-modulus characteristics—the palmar ribs dominate the flexion direction with enhanced bending, while the lateral ribs induce lateral swing motion. The two form a three-dimensional force coupling within the core layer, enabling the finger to achieve fine bending in multiple directions while contracting axially, realistically simulating the combined flexion and lateral swing motion of a real finger. This embedding method enables the two sets of components to form a three-dimensional force coupling at the microscale. The force transmission path is short and direct, ensuring that the contraction and bending forces of the fingers can be efficiently and collaboratively transmitted during the grasping action, thereby improving response speed and output force. The low-modulus ribs are embedded inside the high-modulus core layer, which provides all-round protection during repeated grasping and releasing of the fingers, significantly improving tensile and fatigue resistance and extending service life. By designing the circumferential distribution of the grooves (such as palmar, dorsal, radial, and ulnar sides) as well as their depth and number, the bending stiffness and driving force distribution in different directions can be precisely controlled. This allows the flexion angle, lateral swing amplitude, and movement trajectory of the fingers to be customized according to the target biomimetic needs. For example, lateral ribs can be added to the thumb design to achieve palmar movement, and the palmar rib layout can be optimized in the index finger design to enhance gripping power.

[0172] First-interval arrangement: The first component group and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction.

[0173] Figure 5This is a schematic diagram of a first spaced structure provided for one embodiment of this application.

[0174] In one implementation, reference Figure 5 As shown, in practical implementation, this alternating arrangement structure can be achieved using either stacked assembly or multi-channel co-extrusion processes. The stacked assembly method involves preparing high-modulus material into thin sheet-like strip components as the first component group C1, and preparing low-modulus material into thin sheet-like strip components of the same thickness as the second component group C2. These two types of materials are then stacked alternately in the thickness direction, and the interfaces are fused through hot pressing to form a multi-layered alternating composite structure. The multi-channel co-extrusion method involves designing alternating flow channels within the co-extrusion die, allowing high-modulus and low-modulus materials to alternately converge during extrusion, forming a composite strip component G1 with a multi-layered alternating cross-section in a single process.

[0175] For example, taking SBAS material as an example, high-modulus SBAS material with a modulus of 1.2 MPa is prepared into thin strip-shaped components with a thickness of 0.2 mm and a width of 1 mm as the first component group, and low-modulus SBAS material with a modulus of 0.4 MPa is prepared into thin strip-shaped components of the same size as the second component group. The two are stacked alternately in the thickness direction for 5 layers (forming a structure like...). Figure 5 The high-low-high-low-high arrangement shown in the diagram is hot-pressed at 120℃ and 0.5MPa for 15 minutes to allow the molecular chains at the interlayer interface to diffuse and fuse, resulting in an integrated composite strip-shaped component with a thickness of approximately 1.0mm. When an electric field is applied, the spaced high-modulus components dominate axial contraction, generating a longitudinal contraction driving force; the spaced low-modulus components dominate bending deformation, generating a bending driving force perpendicular to the longitudinal direction. Because the two are alternately distributed in the thickness direction and tightly coupled at the interface, the contraction of each layer of high-modulus components exerts axial constraints on adjacent low-modulus components, while the bending deformation of each layer of low-modulus components induces local deflection of adjacent high-modulus components, causing the entire composite to exhibit uniform composite bending on a macroscopic scale.

[0176] In biomimetic finger applications, this alternating spaced structure serves as the driving force of the finger, with its thickness direction corresponding to the finger's flexion direction. When an electric field is applied, the spaced high-modulus components dominate axial contraction, generating the main driving force for finger flexion; the spaced low-modulus components dominate bending deformation, inducing local bending in adjacent high-modulus components. Because the two are alternately distributed in the thickness direction and tightly coupled at the interface, the contraction of each layer of high-modulus components and the bending of each layer of low-modulus components work synergistically in space, causing the finger to produce a uniform bending curvature along the thickness direction, realistically simulating the smooth flexion movement of a real finger. This alternating spaced structure enables the finger to produce a uniform bending curvature along the thickness direction, realistically simulating the smooth flexion movement of a real finger, and avoiding uneven deformation caused by stress concentration.

[0177] Based on the first type of intermittent composite structure, the alternating arrangement of the two sets of components ensures uniform distribution across the cross-section and symmetrical force transmission path, guaranteeing uniform stress distribution during finger bending and significantly improving the stability and repeatability of gripping movements. Secondly, by adjusting the number of alternating layers (e.g., 3, 5, or 7 layers) and the thickness ratio of each layer, the overall bending stiffness and flexion angle of the finger can be precisely controlled, enabling differentiated design for different phalanges (e.g., the thumb requires greater force, while the index finger requires higher precision). Thirdly, the multi-layered alternating structure forms a composite reinforcement in the thickness direction, effectively suppressing local buckling instability that is prone to occur in single-layer structures, and enhancing the bending stiffness and load-bearing capacity of the finger when gripping heavy objects. Fourthly, by designing an asymmetrical layer distribution on both sides of the thickness direction (e.g., more high-modulus layers on one side and fewer low-modulus layers on the other), unidirectional deflection presets for the finger can be achieved, allowing the finger to naturally flex in a flexed posture without excitation, more closely resembling the physiological curvature of a real finger and simplifying the control strategy.

[0178] In some embodiments, the strip members in the first component group and the strip members in the second component group are arranged in a predetermined spiral path in three-dimensional space, wherein the strip members in the first component group are arranged along the main direction of the spiral to provide axial contraction constraint, and the strip members in the second component group are arranged in a cross spiral to guide torsional motion, so that the composite generates a torsional torque about its axis under electric field excitation.

[0179] Figure 6 This is a schematic diagram of a spiral path arrangement structure provided in one embodiment of this application.

[0180] In one implementation, reference Figure 6 As shown, the spiral path arrangement structure is achieved by arranging two sets of components according to a specific spatial trajectory. In specific fabrication, the strip-shaped component in the high-modulus first component group C1 is used as the core axis and arranged parallel to the main direction (i.e., the axial direction) of the spiral to form the axial skeleton of the composite; the strip-shaped component in the low-modulus second component group C2 is wound around the outer periphery of the first component group in a cross spiral manner.

[0181] The mechanical synergy mechanism of this structure originates from the reconstruction of the driving force direction by the spatial arrangement of the two sets of components. When an electric field excitation is applied, the first set of components, arranged parallel along the axial direction, generates a contraction driving force along the axial direction due to its high modulus characteristics, providing the main constraint for the axial contraction of the composite. The second set of components, arranged in a cross-spiral configuration, also generates contraction along its own length direction due to its low modulus characteristics. However, because of its spiral shape, this contraction force can be decomposed into an axial component along the mandrel direction and a circumferential component perpendicular to the mandrel. The circumferential components generated by the two sets of left- and right-handed spiral symmetrical winding layers are in opposite directions, forming a pair of force couples that apply a torsional moment about the mandrel. At the same time, the axial component of the second set of components and the axial contraction of the first set of components superimpose or constrain each other, jointly regulating the axial strain of the composite. Thus, the axial contraction constraint of the first set of components and the torsional drive of the second set of components naturally synergize in three-dimensional space, enabling the composite to simultaneously generate a combined motion of axial contraction and torsion about the axis under electric field excitation.

[0182] For example, two sets of winding layers can be used, one left-handed and one right-handed, forming a symmetrical cross-spiral network. Key geometric parameters include: helix angle θ (the angle between the winding direction and the axial direction, typically ranging from 30° to 60°), pitch P (the distance between adjacent winding points along the axial direction, usually 2-10 times the component diameter), and cross angle 2θ (the angle between the left-handed and right-handed winding wires). Taking SBAS material as an example, the first component group of strip-shaped members with a diameter of 0.3 mm and a modulus of 1.2 MPa is used as a mandrel and bundled parallel to each other along the axial direction; the second component group of strip-shaped members with a diameter of 0.1 mm and a modulus of 0.4 MPa is wound left-handed with a helix angle of 45° and a pitch of 2 mm, and then wound right-handed with the same parameters to form a cross-spiral coating layer; after winding, a dilute solution of the same material is applied and cured at 100°C, so that the winding layer and the mandrel interface form a chemical bond, obtaining an integrated composite with a helical arrangement structure.

[0183] In some embodiments, the first component group and the second component group are arranged in an antagonistic pair in space, such that when both are excited at the same time, the flexible biomimetic muscle complex produces a rigidity enhancement effect.

[0184] In one embodiment, the antagonistic pair arrangement refers to the spatial arrangement of the first and second component groups in a mutually antagonistic manner. For example, the first and second component groups can be symmetrically arranged on both sides of the bending neutral plane of the composite, or their driving force directions can be geometrically opposite. When only one group is excited, the composite will bend or contract in a single direction; however, when both groups are excited simultaneously, they generate driving forces in opposite directions, forming an antagonistic internal force system within the composite. In biomimetic applications, this antagonistic pair is equivalent to the relationship between agonist and antagonist muscles in biological muscles: the contractile driving force of the first component group attempts to bend the composite to one side, while the contractile driving force of the second component group attempts to bend the composite to the opposite side. The two restrain each other through force transmission within the structure. The result of this internal force antagonism is that the macroscopic bending deformation of the composite is suppressed, but the internal stress is significantly increased, manifested as an increase in the overall stiffness of the composite, that is, an enhanced ability to resist deformation under external forces. This stiffness enhancement effect is actively controllable, and its enhancement magnitude is positively correlated with the excitation intensity of the two component groups.

[0185] For example, taking a bionic finger made of SBAS material as an example, the first component group (modulus 1.2 MPa) is placed on the palmar side of the neutral layer of the finger, and the second component group (modulus 0.4 MPa) is placed on the dorsal side of the neutral layer of the finger. The two are co-cured to form an integrated structure. When only the palmar first component group is excited, the finger produces a flexion movement; when only the dorsal second component group is excited, the finger produces an extension movement. However, when both components are excited simultaneously with the same voltage, the palmar contraction and dorsal contraction antagonize each other, and the macroscopic bending displacement of the finger is almost zero. However, mechanical testing shows that the bending stiffness of the finger is increased to more than three times that of the single-component excitation. For example, when the same load is applied to the fingertip, the deflection is 5 mm without excitation, 4 mm with single-component excitation, and only 1.5 mm with simultaneous excitation of both components, which intuitively demonstrates the stiffness enhancement effect.

[0186] In robotic finger grasping applications, when the finger contacts the target object and needs to maintain a stable grip, the following control strategy can be implemented: After the tactile sensor detects that the contact force has reached a preset threshold, the control system, while maintaining the original grasping drive, applies synchronous excitation to the second component group in the antagonistic pair. At this time, the first component group maintains the flexion force required for grasping, while the second component group generates the opposing extension force. The two form an antagonistic relationship within the finger, significantly improving the overall stiffness of the finger, enabling it to resist the reaction force of the object without deformation or loosening. Taking the grasping of fragile items as an example, this stiffness enhancement mode allows the finger to maintain a gentle contact while possessing sufficient stiffness to support the weight of the object, preventing the object from slipping or deforming due to the finger being too soft. During implementation, the stiffness can be continuously adjusted by adjusting the excitation voltage ratio of the two sets of components to adapt to the grasping requirements of different weights and materials.

[0187] By actively antagonizing the first and second component groups, the stiffness of the composite is actively and controllably adjusted, allowing the same actuator to switch between "flexible compliance" and "rigid support" modes according to task requirements, greatly expanding application scenarios. The stiffness enhancement effect originates from the intrinsic material driving force rather than external constraints, with a fast response speed (millisecond level), and the enhancement amplitude can be precisely controlled by electrical signals. In addition, the antagonistic pair design simulates the synergistic mechanism of the agonist and antagonist muscles of biological muscles, enabling the composite to maintain flexible biomimetic characteristics while possessing the ability to actively change stiffness, overcoming the defect of traditional flexible actuators where stiffness is fixed once formed. In grasping operations, this adjustable stiffness characteristic can significantly improve operational stability and adaptability. For example, when grasping heavy objects, stiffness is increased to prevent deformation, and stiffness is reduced to maintain compliance during fine operations, truly realizing intelligent behavior similar to biological muscles.

[0188] In some embodiments, the strip-shaped component itself has a multi-level structure, including: a first-level structure comprising nanoscale myofibril biomimetic units composed of oriented functional fillers or molecular chains; and a second-level structure comprising micron-level myofibril biomimetic units formed by bundles of multiple myofibril biomimetic units, wherein the micron-level myofibril biomimetic units constitute the basic units of the first component group or the second component group.

[0189] Figure 7 This is a schematic diagram of a multi-level strip structure provided in one embodiment of this application.

[0190] In one embodiment, reference is made to Figure 7As shown, the multi-level structural design of the strip-shaped component deeply simulates the multi-level organizational structure of biological muscle from myofibrils to muscle fibers. The first-level structure is the nanoscale myofibril biomimetic unit P1, composed of oriented functional fillers (such as carbon nanotubes, barium titanate nanowires) or polymer chains. Its core mechanical function is that when an electric field is applied, these nanoscale oriented units generate synergistic micro-contraction along their alignment direction—the functional fillers enhance the local electric field concentration effect through high dielectric constants, while the molecular chains generate entropic elastic contraction through conformational changes. Since all units are oriented in the same direction, their micro-contractions are statistically superimposed, forming a macroscopically visible contractile force along the orientation direction. The second-level structure is the micrometer-level muscle fiber biomimetic unit P2, formed by tens to hundreds of the above-mentioned nanoscale units P1 bundled together in axial parallel. Its mechanical function is to: on the one hand, spatially converge and amplify the micro-contraction force of the nanoscale units, so that the output force of the micrometer-level units reaches the level that can drive macroscopic components; on the other hand, the interfacial shearing action between units in the bundled structure can absorb local stress fluctuations, making the overall contraction smoother and more stable. This progressive mechanical transmission mechanism, characterized by "deformation of nano-units and amplification of convergence of micro-units," enables the driving force to be amplified and integrated stepwise from the molecular scale to the macroscopic scale, allowing strip-shaped components to generate efficient directional drive in a multi-level structural manner similar to biological muscles.

[0191] It should be noted that, in the embodiments of this application, "nanoscale" refers to structural units with characteristic sizes ranging from 1 to 100 nanometers. In actual fabrication, myofibril biomimetic units within this scale range can be obtained by controlling the particle size of the functional filler or the ordered aggregation morphology of the molecular chains. This scale range matches the size of functional proteins within myofibrils in biological muscle and is key to achieving high-precision biomimicry. In the embodiments of this application, "micrometer-scale" refers to structural units with characteristic sizes ranging from 1 to 100 micrometers. In specific implementations, multiple nanoscale myofibril biomimetic units (approximately 10-100 nanometers in diameter) are arranged in parallel along the axis or slightly twisted and bundled to form micrometer-scale myofibril biomimetic units with diameters of approximately 1-50 micrometers. This scale range highly matches the diameter of real muscle fibers in biological muscle (typically 10-100 micrometers), enabling the composite of this application to simulate the hierarchical organization of biological muscle at the microstructural level. By controlling the number of units in the cluster (e.g., clustering 100-1000 nanometer-scale units), the diameter of the micrometer-scale units can be precisely controlled, thereby affecting the mechanical properties and driving response characteristics of the entire composite.

[0192] For example, taking the SBAS triblock copolymer and oriented carbon nanotube composite system as an example, the first-level structure is achieved as follows: carboxylated multi-walled carbon nanotubes (20 nm in diameter and 10 μm in length) are uniformly dispersed in a good solvent of SBAS. Through electrospinning under a high-voltage electric field, the carbon nanotubes are oriented along the fiber axis. Simultaneously, solvent evaporation causes the SBAS molecular chains to self-assemble on the surface of the carbon nanotubes, forming an oriented structure, thus obtaining nanoscale myofibril biomimetic units with a diameter of approximately 100 nm. The second-level structure is achieved as follows: multiple of the above nanofibers are collected and bundled, and then hot-pressed at 120°C, causing the molecular chains at the nanofiber interfaces to diffuse and fuse, forming micron-sized myofibril biomimetic units with a diameter of approximately 10 μm. The carbon nanotubes are highly oriented along the axial direction within the nanofibers, and adjacent nanofibers fuse well at the interface after bundling, forming a continuous force transmission path.

[0193] In the application of high-precision surgical bionic robots, this multi-level structural strip component can serve as the driving unit for minimally invasive surgical instruments. During application, by adjusting the electrospinning process parameters (such as an electric field strength of 15kV, a receiving distance of 15cm, and a solution concentration of 8wt%), the diameter of the nanoscale myofibril bionic unit can be controlled within the range of 80-120nm, ensuring that its microscopic contractile force and response speed match the surgical precision requirements (such as achieving a response time of <50ms). Furthermore, by controlling the number of bundles (such as bundling 500 nanofibers into a single micrometer-sized fiber), the output force of the micrometer-sized myofibril bionic unit reaches the 50mN level, meeting the force required to drive the opening and closing of surgical forceps. At the end of the surgical instrument, multiple micrometer-sized units are grouped according to functional requirements (such as one group for driving clamping and another for driving bending), and an electric field is applied independently via microelectrodes. When fine tissue dissection is required, only a portion of the micrometer-sized units are excited to achieve small-amplitude, high-precision movements; when tissue clamping is required, all micrometer-sized units are simultaneously excited to achieve strong contraction. This hierarchical design allows surgical instruments to possess both the sensitivity for precise operation and the power output capability when necessary.

[0194] This multi-level structure, through its biomimetic design ranging from nanometers to micrometers, allows for the design and control of the actuation behavior of strip-shaped components at multiple scales. The orientation of nanometer-level units determines the anisotropic strength, while the number of bundled micrometer-level units determines the output force, achieving modular design of actuation performance. The microscopic contraction of nanometer-level units is amplified through the progressive convergence of the bundled structure, avoiding stress concentration and failure problems caused by internal defects in traditional single large-size components, significantly improving actuation reliability and fatigue life. This hierarchical structure enables the component to maintain flexibility while possessing high output force density, with a power-to-weight ratio (output force / mass) several times higher than that of traditional flexible actuators, making it particularly suitable for weight-sensitive biomimetic robot applications. This structure highly simulates the multi-level tissue characteristics of biological muscles, enabling the composite to possess mechanical response characteristics similar to those of biological muscles at the microscopic level, providing a material structure foundation for achieving truly biomimetic actuation.

[0195] In some embodiments, the strip member includes reinforcing structures disposed at both ends, the reinforcing structures including at least one of the following: a locally thickened structure, a gradient modulus structure, a fiber reinforcement layer, and an anchoring interface reinforcement layer.

[0196] In some embodiments, in a locally thickened structure, the cross-sectional area of ​​the strip member at both ends is larger than that in the middle region.

[0197] Figure 8 This is a schematic diagram illustrating the cross-sectional area difference setting in one embodiment of this application.

[0198] Among them, reference Figure 8 As shown, in one embodiment, a localized geometrically reinforced zone is formed by increasing material accumulation or cross-sectional dimensions at the ends of the strip-shaped member 100. In specific implementations, this can be achieved through die forming or post-processing: using a variable cross-section extrusion die, the flow rate of the molten material at the ends is slowed down or the die opening is increased during extrusion, forming a variable cross-section strip-shaped member with thicker end regions (L1) and thinner middle region (L2); alternatively, a strip-shaped member with a uniform cross-section is first prepared, and then the same or different materials are added to its end regions through impregnation, spraying, or molding, forming a locally thickened layer after curing. The length of the thickened region is typically 5%-15% of the total length of the member, and the thickening ratio (end cross-sectional area / middle cross-sectional area) is controlled between 1.2 and 3.0 times to ensure stress dispersion without affecting the overall flexibility of the member.

[0199] For example, taking a strip-shaped component made of SBAS material as an example, a variable cross-section extrusion process is used: SBAS melt is passed through a specially designed extrusion die. The gap width in the middle area of ​​the die is 0.5 mm, and the gap width at both ends gradually transitions to 1.0 mm. After extrusion and cooling, a variable cross-section strip-shaped component with a total length of 100 mm, a diameter of 0.8 mm in the middle section, and a diameter that gradually increases to 1.2 mm within a length of 10 mm at each end is obtained. Tensile tests on the prepared component show that, under the same tensile force, the fracture location of the constant cross-section component is randomly distributed, while the fracture location of the variable cross-section component occurs entirely in the middle area, with no fractures at the ends, proving that the end-thickening structure successfully transfers the stress concentration area from the ends to the middle. Further testing of its connection strength with the external skeleton shows that after fixing using the same embedding method, the pull-out force of the variable cross-section component is increased, which is significantly improved compared to the constant cross-section component.

[0200] In biomimetic finger joint actuation applications, the two ends of the strip-shaped component, connecting the fixed end at the base of the finger and the movable end at the tip, are the areas with the most significant stress concentration. During implementation, the two ends of the strip-shaped component with locally thickened structures are fixed to the finger's joint skeleton via mechanical clamping or bonding: the thickened end region is embedded in the groove of the skeleton, and because the cross-sectional size is larger than the groove opening, a self-locking effect is formed; simultaneously, a coupling agent is coated on the surface of the thickened region, co-curing with the skeleton material to form a chemical bond. Taking the thumb as an example, its actuation component needs to withstand alternating loads generated by frequent flexion. The thickened end design ensures that the component remains intact after numerous cycles of actuation, while components without thickening may loosen or break after a certain number of cycles. Furthermore, the thickened region can be pre-drilled with through holes or grooves for inserting fixing pins or binding wires to achieve multiple mechanical fixation.

[0201] The locally thickened structure significantly reduces stress concentration per unit area by increasing the cross-sectional area at the ends. This allows the critical section to shift from the ends to the central region when the component is subjected to tension or bending, preventing premature failure of the ends from causing the entire drive system to fail. The thickened structure provides a larger contact area and a mechanically self-locking foundation for the end connections, improving the connection strength with the external frame and ensuring connection reliability under long-term dynamic loads. This structure is fully integrated into the strip component itself, eliminating the need for additional connectors or reinforcing plates, thus maintaining the overall flexibility and lightweight advantages of the composite. By adjusting the length and thickness ratio of the thickened area, customized designs can be made for different application scenarios and load characteristics. For example, a larger thickness ratio can be used in high-frequency drive applications, while a smaller thickness ratio can be used in space-constrained applications, achieving a balanced optimization between connection reliability and structural compactness.

[0202] In some embodiments, under a gradient modulus structure, the elastic modulus of the strip member is higher in the two end regions than in the middle region, forming a mechanical gradient that increases from the middle to both ends.

[0203] In one embodiment, the material can be prepared using variable temperature treatment, gradient curing, or multilayer composite processes. Variable temperature treatment involves locally heating the two ends of the uniform cross-section strip to a higher temperature (e.g., above the glass transition temperature), causing molecular chain rearrangement or increased cross-linking density in that region, resulting in increased modulus at both ends upon cooling. Gradient curing involves applying additional heat to the two ends or extending the curing time during the material's cross-linking and curing process, resulting in a higher degree of cross-linking than the middle region. Multilayer composite processes involve coating the two ends of the component with one or more layers of high-modulus material through co-extrusion or dip coating, forming a composite structure with a gradual modulus change from the middle to the ends. The length of the gradient transition zone is typically 10%-20% of the total component length, and the modulus change range (modulus at both ends / modulus in the middle) is controlled between 1.5 and 4.0 times to ensure a smooth stress transition without generating new interfacial stress concentrations.

[0204] For example, taking a strip-shaped component made of SBAS material as an example, a variable-temperature treatment process was used to achieve a gradient modulus: A 100mm long, 0.8mm diameter SBAS strip-shaped component with a uniform cross-section was placed in a local heating device with 15mm sections at each end, heated to 130℃ (approximately 30℃ above its glass transition temperature) and held for 10 minutes. Simultaneously, the middle section was kept at room temperature by water cooling. This process caused molecular chain rearrangement and microphase separation in the end sections, strengthening the structure. After cooling, tests showed that the modulus at both ends reached 1.2MPa, while the modulus in the middle section remained at 0.4MPa, with a modulus ratio of 3:1. The modulus in the transition zone (5mm inward from both ends) exhibited a continuous gradient change. Dynamic mechanical analysis of the treated component showed that under cyclic loading at 0.1Hz and 5% strain, the loss factor tanδ of the gradient modulus component was lower than that of the uniform modulus component, indicating a more uniform internal stress distribution. Tensile tests showed that the gradient modulus component exhibited higher fracture strength under end-clamping conditions, effectively improving upon the non-gradient component, and the fracture occurred in the middle section rather than at the ends.

[0205] In the joint actuation application of bionic fingers, the two ends of the strip-shaped component need to be rigidly connected to the fixed end at the base of the finger and the movable end at the tip, which are the areas with the most significant stress concentration. In implementation, the two ends of the strip-shaped component with a gradient modulus structure are fixed to the finger skeleton by adhesive bonding or mechanical clamping: the high-modulus end regions provide sufficient connection stiffness to ensure that the ends do not loosen or creep during rapid finger flexion; the low-modulus middle region maintains the overall flexibility of the component, allowing the finger to produce smooth bending movements. Taking the index finger as an example, its proximal phalanx actuation component needs to withstand frequent flexion and extension cyclic loads. Through gradient modulus design, the attenuation of the connection strength at the end of the component is controlled within a certain range (e.g., 5%) after a large number of cyclic actuations, while components without gradient modulus may experience end creep leading to connection loosening after a small number of cycles. Furthermore, the continuous modulus change in the transition zone allows stress to transition smoothly from the end to the middle, avoiding the creation of new stress concentration points at the junction of the end and the middle.

[0206] The gradient modulus structure, through the increasing modulus distribution from the middle to both ends, makes the stress distribution of the component more uniform when subjected to tensile loads. The high-modulus region at the ends bears the main connection stress, while the low-modulus region in the middle bears the deformation task, achieving synergistic optimization of functional zoning and stress dispersion. Compared with locally thickened structures, gradient modulus can strengthen the ends without changing the geometry of the component, making it more suitable for space-constrained miniaturized applications (such as minimally invasive surgical instruments). The continuous modulus gradient avoids abrupt interfaces, eliminating the risk of interface delamination failure commonly seen in end reinforcement of traditional composite materials, and significantly improving the fatigue life of the component. This structure can be achieved through simple post-processing, without the need for complex molds or equipment, making it cost-effective and easy to mass-produce. The gradient modulus design provides a "flexible to rigid" transition for the end connections of the component, making it more adaptable to connections with external skeletons of different stiffnesses and broadening the application range of the composite.

[0207] In some embodiments, a fiber reinforcement layer is provided, specifically a high-strength fiber reinforcement layer or a mesh constraint layer may be provided around the outer periphery of the two end regions of the strip member.

[0208] Figure 9 This is a schematic diagram of the fiber reinforcement layer configuration provided in one embodiment of this application.

[0209] In one embodiment, the high modulus and high strength properties of fibers can be used to locally reinforce the ends of strip-shaped components. Specifically, this can be achieved through winding, weaving, or prepreg bonding processes. Winding involves impregnating continuous fibers (such as carbon fiber, glass fiber, or aramid fiber) with resin and then spirally winding them at a certain angle (e.g., 45°-90°) at both ends of the strip-shaped component to form a fiber coating layer. After curing, the fiber layer and the component matrix form an integrated structure. Weaving involves inserting pre-woven fiber sleeves into both ends of the component and then using heat shrinking or resin impregnation to ensure the sleeves are tightly bonded to the component surface. Prepreg bonding involves cutting fiber-reinforced prepreg into rings or spiral strips, bonding them to both ends of the component, and then heat-pressing and curing. The thickness of the fiber reinforcement layer is typically controlled at 0.1-0.5 mm, the coating length accounts for 5%-15% of the total component length, and the fiber volume fraction is between 30%-60%, ensuring reinforcement without affecting the overall flexibility of the component.

[0210] For example, taking a strip-shaped component made of SBAS material as an example, local reinforcement was applied to both ends using carbon fiber reinforcement layers: a 100mm long, 0.8mm diameter SBAS strip-shaped component with a uniform cross-section was selected, and epoxy resin adhesive was applied to each end within a 15mm range. Then, two layers of 12K carbon fiber bundles were tightly wound at a 90° winding angle (circumferential winding), with the winding tension controlled at 2N, so that the carbon fiber was tightly adhered to the component surface. After curing at 80°C for 2 hours, the carbon fiber reinforcement layer and the SBAS matrix formed a strong bond. End tensile tests on the reinforced component showed that, under the same clamping conditions, the end failure load of the unreinforced component was 6.5N, and the failure mode was the SBAS material being pulled out from the clamping end; while the end failure load of the carbon fiber reinforced component increased to 15.2N, with a significant increase in end failure load, and the failure mode changed to slow peeling of the fiber reinforcement layer and the SBAS interface, rather than sudden fracture. Scanning electron microscopy showed that the interface between the carbon fiber layer and the SBAS matrix was well bonded, with no obvious debonding or voids.

[0211] In the joint actuation applications of biomimetic robotic arms, the two ends of the strip-shaped component need to be rigidly connected to a metal or composite material skeleton to withstand complex tensile and bending loads. In practice, the two ends of the strip-shaped component with fiber reinforcement are inserted into pre-drilled holes in the skeleton and fixed by mechanical clamping or adhesive bonding. Taking elbow joint actuation as an example, the driving component needs to withstand instantaneous tensile forces up to 20N and bending stresses generated by repeated bending. By locally reinforcing the two ends of the component with carbon fiber reinforcement, the load-bearing capacity at the connection between the end and the skeleton is increased to over 15N, meeting the actuation requirements. Simultaneously, the high stiffness of the fiber reinforcement layer ensures dimensional stability at the end under alternating loads, preventing loosening of the connection due to matrix creep. In actual testing, the strength decay of the end connection of this reinforced component was controlled within a certain range after a large number of cycles of loading, while the unreinforced component showed significant loosening after a certain number of cycles. Furthermore, for scenarios requiring connection to metal joints, threads or grooves can be pre-fabricated on the outer surface of the fiber reinforcement layer to achieve mechanical interlocking with the metal joint.

[0212] By incorporating a fiber-reinforced layer structure and introducing high-modulus, high-strength fiber materials, the load-bearing capacity at the component ends is enhanced, significantly improving the reliability of the connection with the external skeleton. This makes it suitable for driving scenarios with high loads and high impacts. The fiber-reinforced layer and the component matrix form a composite structure. The high elastic modulus of the fibers effectively constrains the lateral deformation of the end region, suppresses the creep flow of the matrix material under clamping force, and extends the connection life. This reinforcement method is completely independent of the material design of the component matrix. The fiber type, winding angle, and number of layers can be flexibly selected according to application requirements to achieve precise control of the reinforcement effect. For example, circumferential winding mainly resists radial expansion, while axial laying enhances tensile strength. While providing end reinforcement, the fiber-reinforced layer does not change the flexibility of the middle region of the component, maintaining the overall driving performance of the component. This structure can also serve as a transition layer for end connections. By adjusting the fiber laying angle, a gradual transition in mechanical properties from a flexible component to a rigid skeleton can be achieved, avoiding stress concentration caused by abrupt changes in stiffness.

[0213] In some embodiments, an anchoring interface reinforcement layer is provided, specifically, an interface reinforcement layer integrally formed with the force transmission structure is provided at both ends of the strip member.

[0214] In one embodiment, the interface between the ends of the strip-shaped component and the external skeleton or connector can be fused and strengthened to form an inseparable integral connection. Specifically, this can be achieved using co-curing, compression molding, or casting processes. Co-curing involves embedding both ends of the strip-shaped component into uncured skeleton material or connector preforms, and then simultaneously cross-linking and curing the component ends with the skeleton material under hot-pressing conditions, with molecular chains at the interface diffusing to form chemical bonds. Compression molding involves placing both ends of the component in a mold cavity, injecting high-strength resin or composite material slurry, and then tightly fusing the injected material with the component ends under high temperature and pressure to form a preset anchoring shape (such as a T-shaped head, barbs, or annular boss). Casting involves placing the component ends in a pre-fabricated skeleton cavity, pouring in a low-viscosity, high-strength adhesive, and curing it. The adhesive penetrates into the micropores or surface texture of the component ends, forming a microscopic mechanical interlock. The thickness and shape of the anchoring interface layer can be designed according to the connection load requirements. Typically, the thickness is 1-3 times the diameter of the component, and the length accounts for 10%-20% of the total length of the component.

[0215] For example, taking the connection between a strip-shaped component made of SBAS material and a carbon fiber composite skeleton as an example, a co-curing process is used to strengthen the anchoring interface: the surface of the SBAS strip-shaped component within a 15mm length at both ends is plasma-treated to increase surface energy, and then pre-embedded in an uncured carbon fiber / epoxy resin prepreg laminate; the entire structure is placed in an autoclave and co-cured for 2 hours at 120℃ and 0.6MPa pressure, allowing interfacial molecular chain diffusion and chemical cross-linking between the SBAS ends and the epoxy resin. Pull-out tests on the connection interface after curing show that the shear strength of the co-cured interface reaches 8.5MPa, with the failure mode being SBAS material fracture rather than interfacial debonding; while the shear strength of the comparative sample using conventional adhesive bonding is only 3.2MPa, with the failure mode being interfacial adhesion failure. Scanning electron microscopy observation shows that at the co-cured interface, SBAS and epoxy resin interpenetrate to form a transition layer of approximately 5μm thickness, with no obvious interface boundary.

[0216] In high-load-bearing actuation applications such as the shoulder joint of biomimetic robots, strip-shaped components need to establish a long-term, reliable connection with a metal or composite material skeleton to withstand complex three-dimensional loads. In practice, the ends of the strip-shaped component with an anchoring interface reinforcement layer are directly integrated into the skeleton as part of the overall design. Taking shoulder joint abduction actuation as an example, the two ends of the SBAS actuation component are pre-embedded in a carbon fiber-reinforced polyetheretherketone (PEEK) skeleton preform. Through compression molding, the skeleton material flows at high temperatures and covers the ends of the component. Simultaneously, stepped or barbed structures are designed in the end regions, allowing the skeleton material to solidify and form a mechanical interlock. The molded skeleton and actuation component form an inseparable whole, maintaining stable connection while withstanding dynamic loads up to 50N during shoulder joint abduction. Furthermore, in scenarios requiring replacement of the actuation component, modular replacement can be achieved through the design of detachable anchoring interfaces. For example, the anchoring interface layer can be machined into an external thread structure that mates with the internal threads on the skeleton, ensuring both connection strength and maintainability.

[0217] By setting an anchoring interface reinforcement layer, the inherent weak interface layer in traditional connection methods is completely eliminated through integrated molding, increasing the connection strength to near or even exceeding the strength of the component itself. This fundamentally solves the failure problem at the connection between the flexible actuator and the rigid skeleton. During co-curing or molding, molecular chain diffusion and chemical cross-linking at the interface form a continuously transitioning gradient interface, avoiding stress concentration caused by abrupt changes in stiffness and enabling smooth load transfer. The structure can be customized to the shape of the anchoring end (such as T-heads, barbs, threads, etc.) according to connection requirements, achieving dual reinforcement of mechanical locking and chemical bonding, adapting to various load conditions such as tension, bending, and torsion. The anchoring interface layer can be constructed simultaneously with the force transmission structure (such as the skeleton, joint) during the molding process, reducing subsequent assembly processes and improving manufacturing efficiency and connection consistency. This reinforcement method is particularly suitable for applications with high requirements for miniaturization and integration (such as minimally invasive surgical instruments and bionic insect robots), achieving high-strength connections without the need for additional connectors, maintaining the lightweight advantage of the composite.

[0218] In some embodiments, the preset force transmission structure between the first component group and the second component group includes a biomimetic arrangement configuration, which includes at least one of a biomimetic weaving configuration, a biomimetic feather configuration, a radial arrangement configuration, or a ring arrangement configuration.

[0219] When the biomimetic arrangement is a biomimetic feather configuration, the first component group is arranged along the functional axis of the composite, and the second component group is arranged at a preset angle to the functional axis. The range of the preset angle is consistent with the physiological range of the feather angle of the muscle fibers of the target pennate or semi-pennate muscle.

[0220] In one embodiment, the biomimetic feather configuration refers to a spatial structure design that simulates the arrangement of pennate muscle fibers in biological muscles. The first component group is arranged parallel to the functional axis of the complex, serving as the main driving unit for axial contraction. The second component group is arranged at a preset angle (feather angle) relative to the functional axis, attached to or interwoven on both sides of the first component group, simulating the arrangement characteristics of oblique muscle fibers in pennate muscles. In a specific implementation, it can be prepared using weaving or lay-up processes: the strip-shaped components of the high-modulus first component group are bundled parallel to each other along the axial direction as the core layer; the strip-shaped components of the low-modulus second component group are inclined relative to the axis at a preset feather angle θ (e.g., 0°-30°), and interwoven with the first component group through weaving to form a "herringbone" or "fishbone" arrangement structure; or, a layer-by-layer lay-up method is used, first laying a layer of the first component group along the axial direction, then laying a layer of the inclined second component group on top, alternating stacking to the desired thickness. The range of the pennate angle θ is determined based on the type of pennate muscle being simulated. For example, θ is 0°-15° when simulating a hemipennate muscle, θ is 15°-30° when simulating a pennate muscle, and multiple inclined layers with different angles can be designed when simulating a multipennate muscle. After the arrangement is completed, the two sets of components are integrated into a single structure through co-curing or interfacial bonding to ensure efficient force transmission.

[0221] For example, taking the fabrication of a biomimetic feather-like structure using SBAS material as an example, the fiber arrangement characteristics of the human biceps brachii (penile muscle) are simulated: 20 strip-shaped components of the first component group (0.3 mm in diameter, 1.2 MPa modulus) are selected and bundled parallel to each other along the functional axis to form a 5 mm wide core layer; 40 strip-shaped components of the second component group (0.2 mm in diameter, 0.4 MPa modulus) are selected and tilted at a 20° feather angle relative to the axis. These are then interwoven with the first component group by hand weaving, with each component of the second component group entering from one side of the core layer, crossing the core layer, and exiting from the other side, forming a symmetrical feather-like arrangement. After weaving, the composite is hot-pressed at 120°C for 15 minutes to allow the interfacial molecular chains to diffuse and fuse, resulting in a biomimetic feather-like composite structure with a thickness of approximately 2 mm and a width of 5 mm. Mechanical testing of this composite shows that its axial contractile strain and output force are both improved; while the comparative sample using the same material but with a simple parallel arrangement does not show the same improvement in axial contractile strain and output force as the biomimetic feather-like composite structure. Microscopic observation shows that the tilted arrangement of the second component group generates lateral compression on the core layer during contraction, making the overall structure more compact and improving the force transmission efficiency.

[0222] In the thumb-driven application of bionic robotic hands, it is necessary to simultaneously achieve strong gripping and precise palmar opposition. A bionic feather-like configuration perfectly matches this requirement. In implementation, the feather-like composite is arranged along the functional axis of the thumb, with its proximal end connected to the hand skeleton and its distal end connected to the thumb's distal phalanx. Based on the biomechanical characteristics of thumb palmar opposition, a feather angle of 15° (simulating the hemipenoid muscle) is designed. Under electric field excitation, the composite generates both axial contraction to drive thumb flexion and lateral force through the inclined second component group, driving the thumb to rotate towards the palm, thus achieving natural palmar opposition. When grasping objects of different shapes, the ratio of axial contraction force to lateral force can be changed by adjusting the excitation voltage ratio applied to the first and second component groups—for example, when grasping a cylinder, the first group of excitations is enhanced to strengthen grip, while when grasping a thin sheet, the second group of excitations is enhanced to optimize the palmar opposition angle. Tests show that the thumb prototype driven by this feather-shaped configuration complex can achieve a 0-90° flexion range and a 0-45° palmar rotation range, with improved maximum gripping force and increased similarity to the human thumb in its motion trajectory.

[0223] By employing this biomimetic feather-like structural design, the second component group is arranged at an angle, increasing the number of driving units per unit volume within a limited space. This results in a higher output force density compared to a parallel arrangement, making it particularly suitable for biomimetic applications with high spatial compactness requirements. The angled second component group generates a lateral force perpendicular to the functional axis during contraction, enabling the composite to achieve multi-degree-of-freedom motion (such as simultaneous contraction and torsion) without adding additional driving units, simplifying system complexity. By adjusting the feather angle, the physiological characteristics of different feathered muscles can be precisely matched, making the mechanical output characteristics (contraction speed, output force, stroke) of the composite highly similar to those of the target biological muscle, thus improving biomimetic realism. The lateral compression of the axial units by the angled units in the feather-like configuration enhances the overall structural density and stability, reduces internal voids, and improves energy transfer efficiency. This structural design allows the composite to bear external loads jointly by units in multiple directions, avoiding stress concentration and significantly improving fatigue resistance.

[0224] In one embodiment, the radial arrangement configuration refers to a spatial structure design that simulates the radial muscle fiber arrangement in biological muscles (such as the pectoralis major and deltoid muscles). Multiple component groups are arranged radially outwards from a central point or central axis. In specific implementations, template-guided molding or multi-channel co-extrusion processes can be used: First, a central anchor point or central axis (which can be made of a high-modulus material) is constructed. The strip-shaped components of the first component group are arranged radially outwards from the center in a straight line, forming a skeletal layer. The strip-shaped components of the second component group are interspersed in the gaps between the first component groups, also radially arranged, but with different divergence angles or layered arrangements. Alternatively, a multi-layered radial structure can be used, with smaller radial angles in the inner layers and larger radial angles in the outer layers, simulating the feather-like radial composite characteristics of muscles. The radial angle (divergence angle) typically ranges from 0° to 60°, and can be determined based on the type of muscle being simulated. For example, 15°-30° is used when simulating the anterior deltoid, and 30°-45° is used when simulating the clavicular portion of the pectoralis major. After the arrangement is completed, the ends of each component group are gathered at the central anchor point in the central area by co-curing or mechanical fixation, while the outer perimeter area is fixed by a ring-shaped constraint layer to form a stable radial structure.

[0225] For example, taking the fabrication of a biomimetic radial configuration using SBAS material as an example, to simulate the fiber arrangement characteristics of the human deltoid muscle: 24 strip-shaped components of the first component group with a diameter of 0.3 mm and a modulus of 1.2 MPa, and 48 strip-shaped components of the second component group with a diameter of 0.2 mm and a modulus of 0.4 MPa are prepared. First, a cylindrical central anchor point with a diameter of 2 mm (made of carbon fiber reinforced SBAS) is constructed. The 24 first component groups are arranged radially outward from the central anchor point at 15° equal intervals until they are fixed at a circumference of 20 mm. Then, the 48 second component groups are divided into two groups of 24, which are arranged at radial angles of 25° and 35° respectively, interspersed in the gaps between the first component groups, forming a three-layer composite structure with different radial angles. After the arrangement is completed, the ends of the components are fused to the anchor point by hot pressing at the central anchor point, and the outer periphery is fixed by an annular SBAS strip. Mechanical tests on the radial composite showed that when all components were simultaneously excited, the overall contractile force was increased, with the contraction direction pointing towards the central anchor point. Furthermore, due to the difference in the direction of the contractile force generated by components at different radial angles, the composite could generate a torsional moment around its central axis while contracting. Microscopic observation revealed that components at different radial angles formed a three-dimensional interwoven network in space, resulting in a rich network of force transmission paths.

[0226] In biomimetic shoulder joint actuation applications, multi-directional and multi-angle compound movements are required, and a radial arrangement configuration perfectly matches this need. During implementation, the central anchor point of the radial complex is fixed to the scapular skeleton, and the peripheral divergent ends are connected to different attachment points on the humerus, simulating the anatomical structure of the anterior, middle, and posterior deltoid muscles. Based on the requirements of different shoulder joint movements such as abduction, adduction, and rotation, the radial angle distribution is designed as follows: 30°-45° for the anterior region (simulating the anterior deltoid), 15°-30° for the lateral region (simulating the middle deltoid), and 30°-45° for the posterior region (simulating the posterior deltoid). When shoulder flexion is required, the components in the anterior region are selectively stimulated, causing them to contract and lift the humerus forward; when abduction is required, the components in the lateral region are stimulated; when compound rotation is required, components with different radial angles are stimulated in a specific sequence and intensity combination, causing the contractile forces in all directions to vectorially synthesize and generate a torsional torque. In actual testing, the prototype shoulder joint driven by this radial complex can achieve a range of motion of 0-120° flexion, 0-110° abduction, and 0-60° internal and external rotation, improving the similarity of the movement trajectory to the human shoulder joint movement.

[0227] By arranging components in a radial configuration, the system creates multi-directional driving force vectors in three-dimensional space, enabling a single composite to achieve multiple degrees of freedom of motion independently or collaboratively. This significantly simplifies the complexity of drive systems for multi-degree-of-freedom joints. The radial structure concentrates the driving forces of all components at the central anchor point, creating a force focusing effect and significantly increasing the output force density. This is particularly suitable for proximal joints such as the shoulder and hip, which require high output force. By adjusting the radial angle distribution in different regions, the system can accurately match the anatomical features and mechanical output characteristics of the target muscle, making the kinematics and dynamics of the composite highly similar to those of biological muscles, thus enhancing biomimicry. The spatial interweaving between components in the radial arrangement forms a natural force coupling network. The contraction of some components affects adjacent components through the interweaving points, producing a synergistic effect similar to that of biological muscles. This structure has good scalability and can achieve accurate simulation of more complex muscle groups (such as multipennate muscles) by increasing the number of radial layers or refining the radial angle partitions.

[0228] In one embodiment, the ring-shaped arrangement refers to a spatial structure design that simulates the fiber arrangement of ring-shaped muscles (such as the orbicularis oculi and orbicularis oris) in biological muscles. The component groups are arranged in concentric rings or spiral rings with a central axis or point as the center. In specific implementations, it can be prepared using layer-by-layer winding or ring-shaped weaving processes: a high-modulus first component group of strip-shaped components is arranged parallel to each other along the axial direction as the core layer, and then low-modulus second component group of strip-shaped components is wound layer by layer around its outer periphery. The winding direction of each layer can be the same or alternating (e.g., one layer left-handed, one layer right-handed), forming a concentric ring structure; or, a ring-shaped weaving machine is used to weave the two groups of components into a tubular or disc-shaped structure according to a preset circumferential density and axial spacing, with the inner ring mainly composed of high-modulus components and the outer ring mainly composed of low-modulus components. The number of annular layers is typically 2-10, with each layer having a thickness of 0.1-0.5 mm. The circumferential winding angle (relative to the circumferential direction) ranges from 0°-30° (pure circumferential) or 30°-60° (helical circumferential). To achieve functional differentiation among different component groups, components with different moduli can be assigned to different annular layers. For example, the inner ring may have the first component group to provide radial support and axial contraction, while the outer ring may have the second component group to provide circumferential contraction and torsional drive. After arrangement, co-curing is used to fuse the interfaces of each annular layer, forming an integrated annular multilayer composite structure.

[0229] Taking the fabrication of a biomimetic ring-shaped structure using SBAS material as an example, simulating the eyelid closure function of the orbicularis oculi muscle in the human eye: A first group of strip-shaped components with a diameter of 0.1 mm and a modulus of 1.2 MPa was selected as the inner ring material, and a second group of strip-shaped components with a diameter of 0.1 mm and a modulus of 0.4 MPa was selected as the outer ring material. A ring-weaving process was used to weave two layers on a 10 mm diameter ring mold: the inner ring was formed by cross-weaving the first group of components at a 45° weaving angle, providing basic support and axial contraction; the outer ring was formed by tightly weaving the second group of components at an 85° weaving angle (close to the circumferential direction), forming the dominant layer for circumferential contraction. After weaving, the mold was placed at 120°C for 10 minutes to fuse the interfaces of each layer. After demolding, a ring-shaped composite with an inner diameter of 10 mm, a wall thickness of 0.8 mm, and a width of 5 mm was obtained. Driven by testing, the annular complex showed that stimulating the outer second component group effectively improved the circumferential contraction rate, annular inner diameter, and circumferential contraction force, thus mimicking the eyelid closure action caused by orbicularis oculi muscle contraction. Stimulating the inner first component group effectively improved the axial contraction rate and annular width, producing a slight downward pressure of the eyelid. When both the inner and outer layers were stimulated simultaneously, the annular body, through the synergistic effect of diameter and width reduction, produced a composite movement similar to natural eyelid closure. High-speed video recording showed that the response time of the annular complex from fully open to fully closed was approximately 80 ms, similar to the blink reflex speed of the human eyelid.

[0230] In biomimetic robot facial expression generation applications, a ring-shaped configuration can be used to drive blinking and squinting movements of artificial eyelids, enhancing the naturalness and realism of robot expressions. In implementation, the ring-shaped composite is embedded in the biomimetic eye socket structure, with its outer edge fixed to the orbital skeleton and its inner edge connected to the movable eyelid skin. Multiple independently controlled ring units are designed according to expression requirements: independent ring-shaped composites are set for the upper and lower eyelids, with each segment distributed along the palpebral fissure arc. When blinking is required, the outer second component group of the upper eyelid ring-shaped composite is stimulated, causing it to contract circumferentially and move the eyelid downwards. Simultaneously, the inner first component group is slightly stimulated to produce axial contraction, causing the eyelid to slightly invert during closure. When expressing surprise, the stimulation is relaxed, allowing the eyelid to fully open. When expressing anger or focus, the ring-shaped composites of both the upper and lower eyelids are stimulated simultaneously, causing the palpebral fissure to narrow, creating a squinting effect. In actual testing, the bionic eyelid can achieve a vertical opening and closing range of 0-12mm, with an adjustable closing speed (0.1-0.5s), and can generate smooth and continuous blinking movements, similar to the blinking curve of a real person. In addition, by adjusting the waveform of the excitation voltage, it can also simulate the slow blinking when fatigued or the rapid blinking reflex when startled.

[0231] Through a ring-shaped structural design and concentric ring layers, the composite structure possesses both circumferential and axial contraction capabilities, accurately simulating the complex movement patterns of ring muscles such as the orbicularis oculi (e.g., concentric contraction and axial inward rotation during eyelid closure), meeting the high-fidelity requirements of facial expression-driven motion. The multi-layered ring structure forms functional zones radially, with the outer layer primarily providing the closing force through circumferential contraction, and the inner layer primarily adjusting morphological details through axial contraction. The synergy between these two layers makes the movement more natural and delicate. By adjusting the circumferential winding angle, number of layers, and material modulus, the contraction speed, output force, and movement stroke can be independently controlled, achieving accurate simulation of different expression intensities (e.g., light blinking and forceful eye closure). The ring-shaped structure has natural hollow channels that can embed sensing optical fibers or lubricating media, providing space for subsequent integration of tactile feedback or self-lubricating functions. The structure is formed using a weaving process, allowing for customized ring curvature based on the orbital anatomy curve, achieving a seamless fit with the bionic skull, enhancing integration and aesthetics.

[0232] In some embodiments, the first component group and / or the second component group are divided into multiple independently replaceable fiber modules. Each fiber module is composed of a bundle of strip-shaped components with a preset elastic modulus value, and the ends of the fiber modules are provided with mechanical interfaces and electrical connection interfaces. The fiber modules are assembled in a detachable manner through the mechanical interfaces, and the electrical connection interfaces are used to provide a driving electric field for the strip-shaped components within the fiber modules and to transmit sensing signals when the strip-shaped components have sensing functions. By selecting fiber modules with different modulus values ​​for combination, or changing the radial arrangement order of the fiber modules on the cross-section of the flexible bionic muscle composite, the modulus spatial distribution of the flexible bionic muscle composite can be adjusted, thereby reconstructing its macroscopic motion output characteristics.

[0233] Figure 10 This is a schematic diagram of the connection structure of a fiber module according to an embodiment of this application.

[0234] In one implementation, reference Figure 10 As shown, fiber module M1 is a basic functional unit that can be independently replaced, formed by bundling multiple strip-shaped components 100 with preset elastic modulus values. It is the physical basis for constituting the first component group or the second component group. Each fiber module is assigned a specific elastic modulus value during preparation. For example, by controlling the tensile ratio, filler content, or crosslinking density, the modulus of all strip-shaped components within the module is made uniform, forming standardized units such as "high modulus module," "medium modulus module," or "low modulus module." The fiber module has standardized mechanical interfaces 1001 and electrical connection interfaces 1002 at both ends. The mechanical interface 1001 can adopt a snap-fit, threaded, or plug-in structure (e.g., ...). Figure 10 As shown), it is used to detachably mount the module onto the assembly frame of the composite; the electrical connection interface 1002 employs a flexible contact or pin array (such as... Figure 10 As shown, the module automatically presses against the bus electrodes within the frame upon insertion, providing a driving electric field for the strip-shaped components within the module, and synchronously transmitting sensing signals when the module contains an integrated driving and sensing component. By combining fiber modules with different modulus values ​​as needed—for example, inserting high-modulus modules into the central region and low-modulus modules into the outer peripheral region, or arranging them radially along the cross-section in order of modulus gradient—the spatial distribution of the composite's modulus can be precisely controlled, thereby reconstructing its macroscopic motion output characteristics and enabling rapid adaptation to different application requirements from the same composite matrix.

[0235] It should be noted that the fiber modules adopt a standardized design, with each module containing a certain number (e.g., 10-50) of strip-shaped components with the same modulus value. Both ends of the modules are equipped with mechanical clips and elastic electrical contacts that match the assembly frame. The composite is composed of multiple fiber modules inserted into the frame, and the modules can be replaced independently. By selecting and combining modules with different modulus values ​​(e.g., high-modulus modules E=10MPa, medium-modulus modules E=5MPa, and low-modulus modules E=1MPa), different modulus spatial distributions can be designed. For example, by concentrating high-modulus modules in the central area and arranging medium and low-modulus modules sequentially outwards, a gradient distribution with a high center and low periphery can be formed, achieving three-dimensional bending; alternating high-modulus and low-modulus modules can enhance torsional motion. This modular and reconfigurable design allows the same composite matrix to quickly adapt to different application scenarios by changing different module combinations, such as switching from a grasping task requiring strong contraction to a manipulation task requiring fine bending.

[0236] For example, taking fiber modules made of SBAS material as an example, modules of three standards—high, medium, and low—are prepared. In preparing the high-modulus module (H-type), 30 strip-shaped components with a draw ratio of 5:1 and a modulus of 1.2 MPa are bundled together, with dovetail mechanical interfaces and 4-pin electrical connectors at the ends. In preparing the medium-modulus module (M-type), 30 components with a draw ratio of 3:1 and a modulus of 0.7 MPa are bundled together. In preparing the low-modulus module (L-type), 30 components with a draw ratio of 1.5:1 and a modulus of 0.3 MPa are bundled together. The nine modules can be inserted into the frame in a 3×3 array. Options include: Option A: H-type modules are inserted in the central area, and L-type modules are inserted around the perimeter, forming a modulus distribution with a high center and low periphery; Option B: The modules are arranged alternately in the order of H, M, L, H, M, L, forming a periodic modulus variation. In the driving tests of the two schemes: Scheme A, under electric field excitation, exhibits gradient strain with strong central contraction and weak peripheral contraction, resulting in a three-dimensional bending of the composite material converging towards the center; Scheme B, due to the alternating modulus distribution, produces a wavy deformation mode. Replacing modules is more time-efficient, eliminating the need for rewiring. This modular system allows the same composite matrix to quickly switch between "strong contraction mode" and "fine bending mode," adapting to different task requirements.

[0237] In multi-task applications of biomimetic robots, the end effector of the robotic arm needs to quickly switch drive characteristics according to the task. For example, it requires high output force when grasping heavy objects, and high-precision compliant motion when assembling precision parts. In implementation, a standardized modular assembly frame is built into the robot's forearm, with multiple sets of fiber modules of different modulus values ​​pre-prepared. When the task instruction is "grab a 5kg weight," the control system automatically selects a high-modulus module (H-type) to insert into the central area of ​​the frame, and a low-modulus module (L-type) to insert into the outer periphery, forming a modulus distribution that is high in the center and low at the periphery. This causes the composite to generate strong axial contraction and output a large gripping force when excited. When the task switches to "assemble a micro-bearing," the system prompts the operator to change the module combination, arranging the H-type, M-type, and L-type modules in a radial gradient order to form a smooth modulus transition. This causes the composite to produce uniform bending when excited, achieving high-precision compliant motion. The module replacement process requires no tools, can be completed by the operator in a short time, and the electrical connection interface is automatically connected, requiring no additional wiring. This reconfigurable drive system enables the same robot to flexibly handle a variety of tasks, from heavy-duty handling to precision assembly, significantly improving equipment utilization.

[0238] By setting up a modular fiber structure, standardized modules can be quickly replaced, allowing a single composite matrix to be reconstructed into multiple modulus spatial distributions. This enables cross-mode switching from "powerful output" to "fine motion," greatly expanding the application scenarios of flexible drive systems. The modular design breaks down the complex composite manufacturing process into two stages: standardized module preparation and rapid assembly. Modules can be mass-produced and uniformly inspected, while the composite can be assembled on-site according to needs, significantly reducing manufacturing costs and cycle time. When some modules experience performance degradation due to long-term use, only the corresponding modules need to be replaced, rather than the entire composite, reducing maintenance costs. The independent replaceability of modules allows for rapid iteration of different modulus ratio schemes during the R&D phase, accelerating the development and testing of new motion modes. The standardized design of the electrical connection interface allows modules to automatically connect to the control system during replacement, supporting plug-and-play functionality without requiring professional personnel. By changing the radial arrangement order of the modules, the modulus gradient curve can be adjusted without changing the number of modules, achieving fine-grained control of the composite motion output characteristics and providing a highly flexible system architecture for soft robots.

[0239] In some embodiments, the multiple strip-shaped components in the flexible biomimetic muscle composite can be arranged in space in at least one of the following forms: parallel arrangement; hexagonal close stacking arrangement; nested arrangement in which a large-diameter strip-shaped component nests one or more small-diameter strip-shaped components; at least a portion of the strip-shaped components remain in a bent state in a bundled state, with a bending angle of 0°-180°; at least a portion of the strip-shaped components remain in a torsional state in a bundled state, with a torsional angle of 0°-360°.

[0240] In one embodiment, the multiple strip-shaped components in the flexible biomimetic muscle composite can be arranged in various spatial configurations according to application requirements, with each configuration having specific structural features and mechanical effects.

[0241] The parallel arrangement is specifically a bundle of all strip components arranged in parallel along the same direction, with the axes of each component parallel to each other and the ends aligned. They are then bonded together or bonded to form an integrated bundle structure. This is the most basic and easiest arrangement method to achieve. For example, by hot-pressing 20 SBAS strip components with a diameter of 0.3 mm into a parallel bundle, a drive unit with high axial contraction force density and a single deformation direction can be obtained. This is suitable for scenarios that require strong linear drive, such as the contraction simulation of bionic muscles.

[0242] The hexagonal close-packed arrangement is specifically a arrangement of strip-shaped components in a hexagonal-dense stacking pattern on the cross-section, similar to a honeycomb structure or fiber bundle arrangement. This arrangement is guided and positioned by a hexagonal grid template, ensuring close contact between components and minimizing void ratio. For example, 61 strip-shaped components with a diameter of 0.2mm are arranged in a hexagonal array (1 in the center, with the number increasing layer by layer on the periphery). After hot pressing, the cross-sectional filling rate can reach over 90%, significantly improving the overall stiffness and output force density of the composite. At the same time, the hexagonal symmetrical structure allows for a more uniform stress distribution when the composite is subjected to external forces, making it suitable for applications requiring high load-bearing capacity.

[0243] Figure 11 This is a schematic diagram of a nested arrangement provided in one embodiment of this application.

[0244] Reference Figure 11 As shown, the nested arrangement specifically uses a large-diameter strip-shaped component as the main body, with one or more small-diameter strip-shaped components nested inside or around it. For example, a high-modulus first component group C1 with a diameter of 0.5 mm can be used as the core layer, with three 0.1 mm diameter micropores machined axially inside it. Then, the material of the low-modulus second component group C2 is injected into the micropores in a molten state and solidified to form a "large tube nested with small wires" nested structure; or multiple small-diameter components can be wound around the outer periphery of the large-diameter component to form a composite layer. This arrangement can integrate more functional units within a limited cross-section, achieving efficient space utilization.

[0245] Figure 12 This is a schematic diagram of a pre-bending shape provided in one embodiment of this application.

[0246] Reference Figure 12 As shown, the pre-bending shape specifically refers to at least a portion of the strip-shaped components being given a preset bending state before or during the bundling process, with the bending angle selectable within the range of 0°-180°. In one implementation, such as... Figure 12The bending component assembly 1201 shown comprises multiple strip-shaped components set into a bent state before being bundled together, and the multiple strip-shaped components in the bent state are bundled together to obtain the bending component assembly 1201 of the target shape. In another combined embodiment, the strip-shaped components can be heat-set by winding them around a mold with a specific curvature to obtain permanent bending memory. Then, the strip-shaped components with bending memory properties are bundled together with other straight strip-shaped components to obtain a component assembly with composite properties. In bionic finger applications, pre-bending the components at the knuckles by 15°-30° allows the composite to naturally flex in a flexed posture without stimulation, more closely resembling the physiological curvature of a real finger and simplifying the control strategy.

[0247] It should be noted that, Figure 12 The two forms shown are for illustrative purposes only. The pre-bending forms provided in this application may also include other bending states, such as wavy bending forms, etc. This application does not limit them.

[0248] Figure 13 This is a schematic diagram of a pre-twist configuration provided in one embodiment of this application.

[0249] The pre-twisted configuration specifically involves at least a portion of the strip-shaped components maintaining a preset torsional state in a bundled state. The torsional angle can be selected within the range of 0°-360°. In one implementation, such as... Figure 13 The torsion component assembly 1301 shown is formed by winding multiple strip-shaped components in a bundled state during the bundling process to impart a preset torsion state, thereby obtaining the torsion component assembly 1301 with the target torsion shape. For example, one end of a strip-shaped component is fixed, and the other end is rotated 180° and then heat-set to acquire torsion memory, before being bundled with other components. When an electric field excitation is applied, the pre-torsion component generates a restoring torque along the preset torsion direction, causing the composite to produce an active torsion motion around its axis. This is suitable for applications requiring rotational drive, such as wrist joint rotation simulation.

[0250] In some embodiments, the multiple strip-shaped components in the flexible biomimetic muscle composite are arranged in space to form at least two component groups with different extension directions, and the component groups are force-coupled through shared anchor points, interlacing connection points or flexible substrates.

[0251] In one embodiment, multiple strip-shaped components are arranged in space to form at least two component groups with different extension directions (such as the first component group extending axially and the second component group extending circumferentially or obliquely). In order to achieve the coordination of driving forces in different directions in space, the component groups are coupled by force in the following three ways.

[0252] The shared anchor point method involves converging the ends of different component groups at a common anchor point, forming an integrated connection through co-curing or mechanical fixation. For example, the ends of an axially arranged first component group and an obliquely arranged second component group are jointly embedded in a prefabricated carbon fiber reinforced anchor block. Under hot-pressing conditions, the molecular chains at the interface between the anchor block material and the component ends diffuse and fuse, achieving force convergence and redistribution. This method concentrates multi-directional driving forces at a single anchor point, forming a vector synthesis point of forces, facilitating subsequent connection and transmission with external components. Furthermore, the anchor point area can be concentrated and reinforced, improving the overall load-bearing capacity.

[0253] The interlacing connection point method involves different component groups interweaving and overlapping in spatial arrangement. At the intersections, local force transmission nodes are formed through interfacial bonding or frictional contact. For example, axial and circumferential components are interlaced using a weaving process, and then hot-pressed at 120°C to allow the molecular chains at the intersections to diffuse and form point-like fusion. This method forms a distributed force transmission network along the length of the components, allowing driving forces from different directions to couple stepwise at multiple intersections, avoiding stress concentration. At the same time, the interlaced structure can enhance the overall structural stability of the composite and suppress relative slippage between components.

[0254] The flexible substrate approach involves embedding or attaching different component groups together in a flexible substrate material, with force transmission achieved through the continuous medium of the substrate. For example, an axially arranged first component group and an obliquely arranged second component group can be placed in liquid silicone rubber, and after curing, the two are indirectly coupled through the elastic substrate. In this approach, the substrate can buffer abrupt changes in stiffness between different component groups, making force transmission smoother; at the same time, the substrate can act as a protective layer to isolate the external environment and provide a carrier for integrating other functional components (such as sensors and wires).

[0255] The three force coupling methods mentioned above can be used individually or in combination to form diverse ways of realizing the "preset force transmission structure", enabling the composite to flexibly design the force transmission path according to application requirements.

[0256] In some embodiments, the strip-shaped components can be cut from a pre-prepared anisotropic sheet. For example, a 0.5 mm thick SBAS anisotropic film (1.2 MPa in the stretching direction and 0.4 MPa in the perpendicular direction) is first prepared by a uniaxial stretching process. Then, using laser cutting or mechanical cutting equipment, strip-shaped components with a width of 0.3 mm and a length of 50 mm are cut at specific angles (such as 0°, 45°, or 90°) to the stretching direction. By changing the cutting direction, strip-shaped components with different principal deformation directions can be obtained. Components cut along the stretching direction mainly shrink axially under electric field excitation, while components cut perpendicular to the stretching direction mainly shrink laterally, and components cut at 45° can produce shear deformation. The advantages of this cutting method are: a large number of strip-shaped components with different orientation angles can be cut from a single anisotropic film, achieving efficient material utilization; at the same time, the cutting process is simple and quick, eliminating the need to prepare components separately for each orientation, significantly improving production efficiency.

[0257] In some embodiments, the anisotropy of the strip-shaped member can also be achieved by setting an external constraint structure. Even if the member itself is an isotropic material, its deformation direction is guided by external constraints. Specific implementations include the following three:

[0258] Fiber-reinforced layer method: High-strength fiber bundles (such as carbon fiber or aramid fiber) are longitudinally bonded to the surface of an isotropic strip-shaped component to form a longitudinal reinforcement layer. For example, on the surface of an isotropic SBAS component with a diameter of 0.5 mm, four carbon fiber filaments with a diameter of 0.05 mm are bonded axially. After curing, a composite component with significantly higher longitudinal stiffness than transverse stiffness is formed. When excited by an electric field, the longitudinal deformation is small due to fiber constraint, while the transverse free deformation is large, and the component exhibits anisotropy with transverse contraction as the main characteristic.

[0259] Mesh constraint layer method: An anisotropic mesh constraint layer is wrapped around the surface of the strip-shaped component. For example, a spirally woven nylon mesh sleeve is used. It is longitudinally stretchable and radially constrained. After being wrapped around the surface of the isotropic component, the radial expansion of the component is suppressed when it is excited by an electric field, while the longitudinal extension is allowed to be free, forming anisotropy with longitudinal deformation as the main feature.

[0260] Discrete constraint element method: Discrete rigid constraint rings are arranged at intervals on the surface of the strip-shaped component. For example, a rigid polyimide ring is inserted into the surface of the SBAS component at 2mm intervals. The constraint rings restrict local radial expansion, but allow free deformation in the area between the rings, so that the whole exhibits directional deformation along the axial direction.

[0261] In some embodiments, the electro-actuating material is selected from at least one of dielectric elastomers, ionomer-metal composites, liquid crystal elastomers, conductive polymers, or composites thereof; or, the strip member is composed of a thermally actuating material, a magnetostrictive material, or a photoactuating material.

[0262] In one embodiment, the electro-actuating material used for the strip-shaped component can be selected in various ways depending on the driving mechanism and performance requirements. Dielectric elastomers (such as acrylate elastomers, silicone rubber, and polyurethane) are suitable for driving scenarios requiring large stroke and high output force due to their high energy density, large deformation, and fast response characteristics; their driving mechanism is based on the Maxwell stress effect. Ionic polymer and metal composite materials (such as Nafion-based composite materials) can produce significant bending deformation at low voltages (1-5V), making them suitable for low-voltage driving and underwater operation scenarios; their driving mechanism is based on volume changes caused by ion migration. Liquid crystal elastomers have both anisotropic and reversible deformation characteristics, and can produce controllable contraction and bending through molecular order and disorder transitions, making them suitable for precision driving applications that require precise control of deformation direction. Conductive polymers (such as polypyrrole and polyaniline) undergo volume changes during electrochemical doping / dedoping, and have the advantages of low driving voltage and good biocompatibility, making them suitable for biomedical microdevices. Furthermore, the aforementioned materials can be used in combination. For example, high-dielectric-constant ceramic fillers can be incorporated into dielectric elastomers to increase the dielectric constant, or conductive fillers can be incorporated into liquid crystal elastomers to achieve electrothermal dual-response actuation. It should be noted that the strip-shaped components are not limited to electro-actuated materials. The strip-shaped components can also be composed of thermally actuated materials (such as shape memory polymers, thermally expanding materials), magnetostrictive materials (such as Terfenol-D composite materials), or photo-actuated materials (such as photoresponsive liquid crystal networks, azobenzene polymers). These materials can also generate active deformation under corresponding thermal, magnetic, and optical field excitation, and can achieve composite motion output through multi-component group design and spatial coordination mechanisms.

[0263] In some embodiments, at least a portion of the strip-shaped members in the second component group are made of a conductive polymer, and their driving voltage is lower than that of the first component group.

[0264] In specific embodiments, conductive polymers (such as polypyrrole, polyaniline, and poly(3,4-ethylenedioxythiophene)) can exhibit significant deformation within a low voltage range of 1-5V due to their electrochemical doping / dedoping mechanisms, while traditional electro-actuating materials such as dielectric elastomers typically require driving voltages of hundreds to thousands of volts. By integrating low-voltage driven conductive polymer components into a second component group and combining them with a high-voltage driven first component group (such as a dielectric elastomer) to form a composite, a graded configuration of driving voltages can be achieved. For example, a strong contractile force is generated by applying a 500V high voltage to the first component group, while a fine bending is generated by applying a 3V low voltage to the conductive polymer components in the second component group. This configuration allows for a high-low voltage hybrid drive system, enabling the composite to be compatible with different types of power sources (such as high-voltage sources and batteries). It simultaneously achieves high-output drive and low-power fine-tuning within a single system. The second component group for low-voltage drive can be directly powered by a portable battery, facilitating deployment in applications sensitive to power supply size, such as wearable devices and micro-robots. The high-low voltage partitioning design avoids the insulation complexity and safety risks associated with high-voltage systems, enhancing the system's practicality and safety. The low-voltage characteristics of the conductive polymer components allow for easy integration with low-voltage control circuits and sensors, facilitating the construction of a fully integrated intelligent drive system.

[0265] In some embodiments, the force transmission structure between component groups includes at least one of interlacing connection points, anchoring interfaces, flexible substrates, friction contact surfaces, covalent cross-linking interfaces, or physical winding structures.

[0266] In one embodiment, the pre-defined force transmission structure between component groups can be achieved through various physical or chemical coupling methods to ensure that the driving forces generated by different component groups can be effectively transmitted and coordinated.

[0267] The interlacing connection point is specifically where two sets of components interweave and intertwine in spatial arrangement. At the intersection, local force transmission nodes are formed by hot pressing or bonding. For example, the first set of axially arranged components and the second set of spirally arranged components are interlaced according to the weaving process, and hot pressing at 120°C causes the molecular chains at the intersection interface to diffuse into each other, forming a distributed force transmission network.

[0268] The anchoring interface is specifically achieved by connecting the ends of two sets of components through a shared anchor point. For example, the ends of the first and second component sets are jointly embedded in a prefabricated carbon fiber reinforced anchor block, and an integrated anchoring area is formed through co-curing, thereby realizing the convergence and redistribution of multi-directional driving forces at the ends.

[0269] A flexible substrate is specifically composed of two sets of components embedded or attached to a flexible substrate material (such as silicone rubber or polyurethane film). Indirect force transmission is achieved through the continuous medium of the substrate, meaning that the substrate can buffer sudden changes in stiffness and make force transmission smoother.

[0270] The friction contact surface is specifically formed when two sets of components generate static friction at the interface through pre-pressure or interference fit, and rely on friction to transmit tangential force. For example, the second component group is spirally wound around the outer periphery of the first component group with a certain tension. The positive pressure generated by the winding tension makes the interface friction force sufficient to transmit the driving force.

[0271] A covalent cross-linked interface is specifically formed when two sets of components are bonded together by a chemical cross-linking reaction at the interface. For example, after coating the surface of the first set of components with an uncured slurry of the same material, the second set of components is attached. During the co-curing process, the molecular chains at the interface undergo chemical cross-linking to form an inseparable integrated interface.

[0272] The physical winding structure specifically involves two sets of components achieving force coupling through mechanical winding. For example, the second component group is wound around the outer periphery of the first component group in a cross-spiral manner, and forces are transmitted between the winding layers and between the winding layers and the core layer through geometric interlocking.

[0273] The aforementioned force transmission structures can be used individually or in combination according to application requirements, forming diverse ways to realize the "pre-set force transmission structure" between component groups. This allows the composite to flexibly design force transmission paths for different motion modes, ensuring effective coordination between contraction driving force and bending driving force in space.

[0274] In some embodiments, the flexible biomimetic muscle composite further includes a flexible isolation layer located between multiple strip-shaped members and / or between groups of members. The flexible isolation layer is a porous or mesh structure used to reduce frictional losses between adjacent strip-shaped members and / or to serve as a channel for containing and transporting functional fluids.

[0275] In one embodiment, the flexible isolation layer employs a porous or mesh-like structure to reduce frictional losses between adjacent strip components during dynamic driving and / or serve as a channel for containing and transporting functional fluids. Specifically, before the strip components are bundled together, a layer of porous flexible material (such as foamed silicone rubber or electrospun fiber membrane) is coated onto the surface of the components using dip coating, spraying, or co-extrusion processes. Alternatively, a pre-fabricated mesh-like isolation layer (such as nylon mesh or PTFE mesh fabric) is laid between the component groups, followed by hot pressing to create a stable physical bond between the isolation layer and the component interface. The pore size of the porous structure can be controlled within the range of 10-100 μm, and the porosity can be set within the range of 30%-70%, maintaining flexibility while providing fluid containment space. The mesh-like structure balances the isolation effect with force transmission efficiency by adjusting the mesh density and filament diameter (e.g., 50-200 mesh).

[0276] Figure 14 This is a schematic diagram of a flexible isolation layer provided in one embodiment of this application.

[0277] Reference Figure 14As shown, a flexible isolation layer 1401 with a porous or mesh structure forms a physical isolation between adjacent components, avoiding direct contact friction between component surfaces, significantly reducing wear and energy loss during dynamic driving, and extending the service life of the composite. The porous structure can serve as a channel for containing and transporting functional fluids (such as lubricants, coolants, and dielectric enhancement fluids), achieving self-lubrication or active cooling through capillary action or external pressure, thereby improving driving efficiency and thermal management capabilities. The presence of the isolation layer allows for small gaps between components, enabling relative slippage when the components are bent, avoiding local stress concentration caused by rigid constraints, and improving the flexibility and deformation capacity of the composite. The mesh structure provides isolation while still allowing some force to be transmitted through the mesh nodes, achieving a balanced design of "isolation and force transmission". This structure can also serve as a carrier for a sensing functional layer. For example, by filling porous materials with ionic liquids or conductive slurries, a distributed sensing network can be constructed to achieve real-time deformation monitoring.

[0278] Example of a flexible isolation layer between strip-shaped components:

[0279] Taking SBAS strip component bundles as an example, before the components are bundled, an electrospinning technique is used to coat the surface of each 0.3mm diameter SBAS component with a 20μm thick polyvinylidene fluoride nanofiber membrane (pore size of about 5μm, porosity of 75%). Then, 100 coated components are bundled in parallel and lightly hot-pressed at 80℃ to make the fiber membranes entangle with each other without destroying the porous structure. The lubricant is evenly distributed in the porous network through capillary action, achieving long-term self-lubrication.

[0280] Example of a flexible isolation layer between component groups:

[0281] In a biomimetic finger composite consisting of a first component group (arranged axially parallel) and a second component group (spirally wrapped), a layer of polytetrafluoroethylene (PTFE) mesh (100 mesh, 50 μm filament diameter) is laid between the two groups. During hot pressing, the mesh forms a mechanical interlock with the interfaces of the two components, but does not chemically bond, thus preserving the interface slippage capability. In dynamic finger flexion tests, the mesh isolation layer reduces the frictional resistance between the first and second component groups, while the mesh nodes can still transmit some tangential force, ensuring effective coordination between the two drives. When active cooling is required, a dielectric coolant is injected into the mesh layer via a micro-pump. The liquid is evenly distributed along the mesh channels, carrying away the heat generated by the drive and reducing the temperature rise of the composite during continuous operation.

[0282] In some embodiments, the flexible biomimetic muscle composite also includes a flexible strain sensor integrated thereon for real-time monitoring of the composite’s deformation and transmitting feedback signals to the control system.

[0283] In specific embodiments, the flexible strain sensor can be made of resistive, capacitive, or piezoelectric flexible sensing materials. For example, conductive nanocomposite materials (such as carbon nanotube / silicone rubber composites) can be directly prepared on the surface of the composite through screen printing or coating processes to form a sensitive layer conformally bonded to the composite; or a pre-fabricated flexible strain sensing film (such as a polyvinylidene fluoride piezoelectric film) can be integrated into key parts of the composite (such as the ends or bending neutral layers) through bonding or co-curing. The sensor deforms synchronously with the deformation of the composite, converting the deformation into changes in resistance, capacitance, or voltage signals, which are then transmitted to an external control system via wires or a wireless module. This system enables real-time monitoring of the composite's motion state, providing feedback signals for closed-loop control. This allows the system to adjust drive parameters based on actual deformation, improving control accuracy and response speed. The flexible sensors are matched to the composite's mechanical properties, without affecting its compliance or drive performance. By arranging sensors at multiple points, the system can acquire spatial deformation distribution information of the composite, providing a data foundation for motion pattern recognition and fault diagnosis. This integration method upgrades the composite from a simple actuator to an intelligent drive unit with sensing capabilities, laying the foundation for autonomous control and human-machine interaction in soft robots.

[0284] In some embodiments, at least a portion of the multiple strip-shaped components are integrated drive-sensing components, which specifically include at least one of the following: a first drive-sensing component, composed of a piezoelectric material and an electro-actuating material, configured to acquire piezoelectric signals generated by deformation in real time while undergoing active deformation under electric field excitation; and a second drive-sensing component, composed of an ion-type electro-actuating material, and having an electrode array disposed inside or on its surface, configured to monitor changes in ion concentration distribution representing the deformation state of the component in real time when undergoing active deformation under electric field excitation.

[0285] In a specific embodiment, the integrated drive-sensing component includes at least one of the following types: A first drive-sensing component is composed of a piezoelectric material and an electro-actuating material (e.g., lead zirconate titanate piezoelectric ceramic nanoparticles dispersed in a dielectric elastomer matrix, or a layered composite structure). It is configured to acquire the electrical signal generated by the deformation in real time using the piezoelectric effect while undergoing active deformation under electric field excitation. The greater the deformation, the stronger the piezoelectric signal, and the signal corresponds to the deformation amount. A second drive-sensing component is composed of an ionic electro-actuating material (e.g., an ion polymer and a metal composite material), and has a microelectrode array inside or on its surface. It is configured to monitor the change in ion concentration distribution inside the component in real time when undergoing active deformation under electric field excitation. The concentration distribution caused by ion migration is directly related to the deformation state, and the deformation information can be inferred from the potential or current changes acquired through the electrode array. Unlike the aforementioned externally added flexible strain sensors, the sensing function of the integrated drive-sensing component originates from the intrinsic properties of the material or the component's own structure, requiring no additional sensing elements. It enables deep integration of sensing and driving functions at the material level, giving the component itself "self-sensing" capabilities, resulting in a more compact structure and higher integration. The sensing signal originates directly from the physicochemical changes during the driving process, with fast response and no lag, enabling instantaneous monitoring of deformation. Since the sensing unit and the driving unit are the same entity, it avoids problems such as interface debonding and signal crosstalk that may exist with external sensors, significantly improving reliability. This integrated design allows each strip component to independently sense its own deformation, providing the possibility for distributed sensing and fine-grained control, and is particularly suitable for miniaturized biomimetic applications with strict space and weight constraints.

[0286] In some embodiments, under electric field excitation, the ratio of the main deformation along the longitudinal extension direction to the deformation along the transverse direction of the strip member is greater than 5:1.

[0287] In specific implementations, this ratio (deformation ratio greater than 5:1) is achieved through the aforementioned anisotropic fabrication processes (such as molecular chain orientation, directional arrangement of functional fillers, or external constraint structures). For example, SBAS components with highly oriented molecular chains, achieved through high-stretching, exhibit significant contraction or elongation deformation along the orientation direction (longitudinal) under electric field excitation, while deformation perpendicular to the orientation direction (lateral) is significantly suppressed due to molecular chain constraint. This high ratio ensures that the deformation of the strip-shaped component under electric field excitation is highly concentrated in the preset direction, providing a clear directional basis for the spatial coordination of subsequent multi-component groups. The significant suppression of lateral deformation avoids energy dissipation in unexpected directions during the driving process, improving energy conversion efficiency. The high ratio ensures that the deformation directionality of each component does not interfere with each other after the components are bundled into a composite, facilitating the precise design and control of composite movements such as contraction, bending, and torsion. This quantitative indicator provides a clear basis for standardized testing and quality control of component performance, ensuring that different batches of components can work stably and collaboratively in the composite.

[0288] In some embodiments, the first component group and the second component group are controlled by independent electrode lines to achieve independent addressing and driving of the first component group and the second component group.

[0289] In specific implementations, each component group can receive independent electric field excitation by configuring a dedicated electrode path for each group. Specifically, during the composite fabrication process, the electrodes of all strip-shaped components in the first component group can be connected in parallel to a first bus electrode, and the electrodes of all strip-shaped components in the second component group can be connected in parallel to a second bus electrode. The two bus electrodes are insulated from each other and are respectively connected to independent output channels of an external controller. Alternatively, when the component group contains multiple independently controllable sub-units, more detailed independent electrode lines can be set for each sub-unit. With this independent electrode layout, the controller can apply electrical signals with different voltage amplitudes, waveforms, frequencies, phases, or durations to the first and second component groups, thereby achieving independent adjustment of the excitation parameters of the two groups of components. The dedicated electrode pathway allows the composite to flexibly adjust the driving force ratio between the first and second component groups according to motion requirements. For example, it can enhance the excitation of the first group when strong contraction is needed, and enhance the excitation of the second group when fine bending is needed. It provides the hardware foundation for time coordination mechanisms (such as different start times, different durations, and different frequencies), enabling the composite to generate complex dynamic motions such as oscillation, waves, and spirals. Independent addressing control allows the same composite to quickly switch between different motion modes without changing the hardware structure, greatly improving the system's adaptability and programmability. This control method can be extended to multi-component group systems (such as those containing a third component group) to achieve more complex multi-channel collaborative control and meet the needs of advanced biomimetic motion.

[0290] In some embodiments, the second component group is configured to generate axial contraction deformation under electric field excitation to assist the first component group in achieving axial contraction; the first component group and the second component group are controlled by independent electrode circuits respectively, and in response to electric field excitation with different start times and / or different durations, the contraction deformation of the first component group and the second component group forms a preset sequential or superimposed relationship in time, thereby achieving smooth establishment and attenuation of contraction force during the axial contraction process of the composite.

[0291] In some specific embodiments, the second deformation generated by the second component group under electric field excitation may also include contractile motion. The second component group is configured to generate axial contraction rather than bending deformation, and a single axial contractile motion is achieved through timing control. Specifically, the first and second component groups are arranged parallel along the same axis, and both groups are controlled by independent electrode circuits. Taking the biceps drive of a bionic arm as an example, when a smooth elbow flexion movement is required, the control system first applies a low voltage excitation (e.g., 300V, lasting 50ms) to the second component group, causing it to produce a weak pre-contraction, achieving "compliant initiation" of the muscle; then, a main excitation voltage (600V, lasting 300ms) is applied to the first component group, causing it to produce a strong axial contraction, dominating the main phase of the elbow flexion movement; finally, the excitation of the first component group is cut off 50ms before the end of the movement, while the excitation of the second component group is maintained to slowly decay, so that the contractile force gradually weakens, achieving "smooth release". By adjusting the start time difference (e.g., 10-200ms) and duration ratio of the two sets of excitations, the upward slope, peak duration, and downward slope of the contractile force can be precisely controlled. Unlike complex movements such as oscillations and waves generated through time coordination, this control focuses on optimizing the dynamic characteristics of single-axis contraction, thus transforming the originally rigid "on / off" contraction into a controllable gradual process, simulating the force control characteristics of biological muscle contraction. This approach is particularly suitable for scenarios requiring precise force control, such as when a bionic hand grasps fragile items like eggs, smoothly establishing contractile force to avoid impact damage and smoothly decaying contractile force to achieve gentle release.

[0292] The above is a description of the structural features of the flexible biomimetic muscle composite provided in the embodiments of this application.

[0293] The following section provides a detailed explanation of the driving functions achievable by the flexible biomimetic muscle complex.

[0294] The first component group and the second component group are coupled through a preset force transmission structure, so that when the electric field excitation is applied, the first driving force generated by the first component group, including the first deformation, and the second driving force generated by the second component group, including the second deformation, coordinate with each other in space to realize the motion output of the flexible biomimetic muscle complex that is consistent with the movement characteristics of biological muscles.

[0295] Based on any embodiment of the flexible biomimetic muscle complex provided in this application, the difference in axial stiffness between the first component group and the second component group causes the first component group to undergo a first deformation under electric field excitation. The first deformation includes at least contraction or elongation motion, and causes the second component group to undergo a first contraction deformation or a first elongation deformation along its axial direction, or a second contraction deformation or a second elongation deformation, under electric field excitation.

[0296] Under the influence of the difference in axial stiffness between the first and second component groups, they can produce axial deformations of different amplitudes or directions under electric field excitation, thereby achieving fine differentiation of driving functions. Specifically, when the two component groups use the same material (same elastic modulus E) and have the same length L, the difference in axial stiffness depends entirely on the difference in cross-sectional area A. That is, the first component group, with higher axial stiffness (i.e., a larger product of EA), produces a smaller amplitude of axial deformation (contraction or elongation) under the same electric field excitation, but can withstand and transmit a larger driving force, thus making it suitable as the main driving unit. The second component group, with lower axial stiffness, produces a larger amplitude of axial deformation due to a smaller product of EA, but has a smaller output force, making it suitable as an auxiliary adjustment unit. When two sets of components use different materials or lengths, the combined effect of axial stiffness differences makes deformation distribution more flexible: for example, the first component group uses a high-modulus material and a short-size design to achieve high stiffness, generating a small-amplitude, high-output force main contraction; the second component group uses a low-modulus material and a long-size design to achieve low stiffness, generating a large-amplitude, low-output force auxiliary contraction. Both can work together to smoothly establish, maintain, and decay the contraction force through independent electrode control. When the two sets of components use opposite polarity drives, antagonistic movements of one group contracting and the other elongating can also be achieved, for example, bidirectional drive in a bionic joint. This implementation directly links axial stiffness differences with deformation amplitude, internalizing the role division of the drive unit (main drive and auxiliary adjustment) into the material and structural design, eliminating the need for complex control; by adjusting the two sets of excitation voltages through independent electrodes, the total output force and deformation rate can be continuously controlled; the combination of contraction and elongation provides the structural basis for bidirectional drive and active reset, eliminating the need for additional passive components such as springs. For example, taking the proximal phalanx actuation of a bionic finger as an example, the first component group consists of 10 parallel bundles of SBAS high-stiffness components with a diameter of 0.4 mm, a modulus of 1.2 MPa, and a length of 50 mm. The second component group consists of 10 parallel bundles of low-stiffness components with a diameter of 0.2 mm, a modulus of 0.4 MPa, and a length of 50 mm. When a voltage of 600 V is applied to both groups simultaneously, the first component group generates a 3% primary contraction (output force 2.5 N), and the second component group generates an auxiliary contraction (output force 0.5 N). Together, they cause the finger to flex at 45° and achieve a gripping force of 3.0 N. When a precise release of an object is required, the excitation of the first component group is maintained while the excitation of the second component group is reduced, weakening the auxiliary contraction and smoothly reducing the total contraction force to 1.0 N, achieving a gentle release. If the electrodes of the second component group are reversed, causing it to elongate, active extension and reset of the finger can be achieved, with a reset speed that is improved compared to the passive spring reset speed.

[0297] In some embodiments, the second component group is configured to generate a first contraction deformation along the axial direction under electric field excitation, and to cooperate with the first deformation to complete motion output.

[0298] In a specific implementation, the second component group is configured to generate a first axial contraction deformation under electric field excitation. The core of the motion output, which works in conjunction with the first deformation of the first component group, lies in the difference in axial stiffness and the superposition effect of the mechanical forces between the two component groups. Specifically, the first component group adopts a high axial stiffness design (e.g., large cross-sectional diameter or high elastic modulus), generating a main contraction deformation with a small amplitude but large output force under electric field excitation, serving as the primary power source for motion. The second component group adopts a low axial stiffness design (e.g., small cross-sectional diameter or low elastic modulus), generating an auxiliary contraction deformation with a large amplitude but small output force under electric field excitation, serving as a supplementary power source for motion. The two component groups are arranged parallel along the same axial direction and mechanically coupled through a shared anchoring interface. When electric field excitation is applied simultaneously, the high-stiffness contraction of the first component group provides the basic driving force, while the low-stiffness contraction of the second component group incrementally supplements the total displacement through its large-amplitude deformation. The linear superposition of the contractions makes the total contraction amplitude of the composite greater than that of a single excitation, and the output force is accumulated due to the parallel connection of the two groups. Taking the flexion actuation of a bionic finger as an example, the first component group is composed of a bundle of high-stiffness components, generating primary contraction (output force 3N) under 600V voltage. The second component group is composed of a bundle of low-stiffness components, generating auxiliary contraction (output force 1N) under the same voltage. The two components work together to achieve a finger output force of 4N, realizing a large stroke and high force output that a single component group cannot achieve. This collaborative mechanism naturally allocates primary and auxiliary roles through stiffness differences, allowing the composite to obtain a larger deformation range while maintaining high output force. The mechanical superposition of the two sets of contraction deformations does not require complex control, and the structural design itself guarantees the collaborative effect. This design provides a simple and effective solution for applications that require both large stroke and high force output (such as the leg drive of jumping robots).

[0299] In some embodiments, the second component group is configured to generate a second axial contraction deformation under electric field excitation, which spatially counteracts the first deformation of the first component group and works together to complete motion output.

[0300] In a specific implementation, the second component group is configured to generate a second axial contraction deformation under electric field excitation, which spatially counteracts the first deformation of the first component group, collaboratively completing the motion output. Specifically, the two component groups generate axial forces in opposite directions and form internal force antagonism through a shared anchoring point. The specific implementation is as follows: the first and second component groups are arranged parallel along the same axis, with one end of each anchored to the same skeletal point (e.g., the proximal end of a joint), and the other ends connected to opposite sides of the moving component (e.g., the flexion and extension sides of a joint). When only the first component group is excited, its axial contraction drives the moving component to deflect to one side; when only the second component group is excited, its axial contraction drives the moving component to deflect to the opposite side; when both component groups are excited simultaneously, they generate contraction forces in opposite directions, forming internal force antagonism at the shared anchoring point, causing the macroscopic displacements of the moving component to cancel each other out or partially cancel each other out, but significantly increasing the internal stress, thereby achieving active adjustment of joint stiffness. By adjusting the ratio of the two sets of excitation voltages, the joint's resistance to external forces can be continuously adjusted while maintaining a relatively constant joint angle. For example, increasing the excitation of the first component group while decreasing the excitation of the second component group allows the joint to maintain a stable angle while being preloaded to one side, adapting to loads in different directions. This collaborative mechanism is the first to realize the biomimetic functions of agonist and antagonist muscles in artificial muscles, enabling a single joint to have both motion drive and stiffness adjustment capabilities. The stiffness enhancement effect generated by internal force antagonism can replace the independently set brakes or variable stiffness mechanisms in traditional robots, significantly simplifying the system structure. The stiffness matching design of the two sets of components allows the joint to maintain compliance (single-set excitation) and switch to rigid mode as needed (dual-set excitation) during dynamic movement, providing biomimetic robots with motion control characteristics closer to those of biological muscles.

[0301] In this application simulating the coordinated forearm flexion and extension driven by the biceps and triceps brachii muscles, this embodiment achieves antagonistic function through two sets of components with similar axial stiffness. The first component set simulates the biceps brachii, consisting of 15 parallel bundles of SBAS components with a diameter of 0.35 mm and a modulus of 0.8 MPa, with one end anchored to the scapular skeleton and the other end connected to the proximal radius. The second component set simulates the triceps brachii, consisting of 15 parallel bundles of the same material and diameter, sharing the scapular anchor point with the first component set, and the other end connected to the proximal ulna (elbow extension side). The axial stiffness of the two sets of components is similar (EA product is 0.8 × 0.096 ≈ 0.077), ensuring effective internal force antagonism when simultaneously stimulated. When elbow flexion is required, a 600V voltage is applied to the first component group alone, causing its axial contraction to pull the radius and flex the forearm. When elbow extension is required, a 600V voltage is applied to the second component group alone, causing its axial contraction to pull the ulna and extend the forearm. When lifting heavy objects and maintaining joint stability, a 400V voltage is applied to both component groups simultaneously. The two groups generate opposing contractile forces at the elbow joint, creating internal force antagonism. This allows the forearm to maintain a 45° flexion angle while significantly increasing joint stiffness compared to single-group excitation, effectively resisting the downward torque generated by the heavy object. During the eccentric contraction phase when slowly lowering the heavy object, while maintaining a 500V excitation on the first component group, the excitation on the second component group is gradually increased from 200V to 500V, gradually enhancing the antagonistic effect. The forearm descends smoothly under resistance, and the descent speed can be precisely controlled by adjusting the ratio of the two voltage groups. This application achieves biomimetic antagonistic function between agonist and antagonist muscles in artificial muscles. By simultaneously exciting two sets of components with similar stiffness to generate internal force resistance, a single joint can simultaneously possess the dual capabilities of bidirectional drive and active stiffness adjustment. When grasping heavy objects, by adjusting the ratio of the two sets of excitation voltages, the stiffness resisting external forces can be continuously adjusted while keeping the joint angle constant. Compared with the traditional solution that requires a separate brake, the weight is reduced and the structural complexity is decreased. The precise control of eccentric contraction enables the robot to simulate the natural action of the human body slowly lowering heavy objects, and the energy recovery efficiency is improved compared with passive damping.

[0302] In some embodiments, the first component group and the second component group are controlled by independent electrode circuits. In response to electric field excitation with different start times and / or different durations, the contraction deformation of the first component group and the second component group forms a preset sequential or superimposed relationship in time, thereby achieving smooth establishment and attenuation of contraction force during the axial contraction process of the composite.

[0303] In a specific implementation, based on the arrangement of two sets of components in a mutually antagonistic manner (i.e., the first and second component sets are connected to opposite sides of the joint movement direction, and their contraction directions are opposite), the excitation timing is controlled by independent electrode circuits, enabling smooth establishment and attenuation of contractile force during axial contraction. Specifically, taking elbow flexion as an example, the first component set simulates the biceps brachii (flexion side), and the second component set simulates the triceps brachii (extension side). When a smooth initiation of flexion from extension is required, a 300V voltage is first applied to the second component set for 50ms, causing a slight contraction. Specifically, since the second component set is connected to the extension side, its contraction generates a slight extension pretension at the joint, eliminating the joint gap and establishing initial force balance. After a 50ms delay, a 600V voltage is applied to the first component set, causing a major contraction. At this point, the flexion force of the first component set must overcome the pretension of the second component set to drive joint flexion. This "antagonistic breakthrough" process makes the establishment of flexion force no longer steep, but rather presents a smooth upward slope. When flexion is in place and posture needs to be maintained, two sets of excitations (500V for the first set and 200V for the second set) are maintained simultaneously. The difference in contractile forces between the two sets is used to maintain the joint angle, and the joint stiffness is significantly improved due to the antagonism of internal forces. When slow extension is required, the excitation of the first set of components remains unchanged, while the excitation of the second set of components is linearly increased from 200V to 500V, so that the extension force gradually increases and tends to balance with the flexion force, and the joint smoothly transitions to a neutral position in the process of resistance. By continuing to increase the excitation of the second set of components and decreasing the excitation of the first set of components, smooth movement towards the extension side can be achieved. By adjusting the onset time difference (10-200ms) and duration ratio of the two sets of excitations, the rising slope, peak duration, and falling slope of the contractile force can be independently controlled—the larger the onset time difference, the longer the pretension build-up time, and the gentler the flexion initiation; the duration ratio determines the balance point of the antagonism of internal forces, thus affecting the steady-state stiffness of the joint. The beneficial effects of this timing coordination mechanism are: combining the structural characteristics of "mutual antagonism" with the control of the time dimension, enabling the joint to have active variable stiffness and stepless speed regulation capabilities during movement; eliminating mechanical backlash through pretension, making the start of movement smoother and avoiding the impact caused by gear backlash or slack in traditional switch-type drives; and achieving deceleration at the end of the movement through the gradual dissipation of antagonistic forces, without relying on external brakes, thus improving energy efficiency.

[0304] In some embodiments, the first deformation generated by the first component group includes contraction or elongation, and the second deformation generated by the second component group includes at least bending. The first deformation and the second deformation work together to generate torsional, spiral, oscillating, or wave motion.

[0305] In a specific implementation, the coordinated torsional motion generated by the first deformation (contraction or elongation) and the second deformation (bending) is based on spatial geometric constraints and the principle of force vector synthesis. When the first component group contracts or elongates along the axial direction, its driving force is transmitted along the axial direction; when the second component group undergoes bending deformation, its driving force is perpendicular to the axial direction or at a certain angle to the axial direction. By coupling the second component group with the first component group in a specific spatial configuration (such as spiral wrapping, eccentric arrangement, or cross arrangement), the bending driving force of the second component group can be decomposed into axial and circumferential components, wherein the circumferential component applies a torsional moment about the axis to the first component group. For example, when the second component group wraps around the outer periphery of the first component group in a spiral manner, its bending deformation generates a driving force along the tangent direction of the spiral line. This force can be decomposed into a component parallel to the axial direction of the first component group and a circumferential component perpendicular to the axial direction. The circumferential component causes the first component group to rotate about the axis. When two sets of second component groups with left-right helical symmetry are excited simultaneously, their circumferential components form a couple, further enhancing the torsional effect. This collaborative mechanism requires only two sets of components to achieve a combined contraction (or elongation) and torsion motion, eliminating the need for additional torsion drive units and significantly simplifying the system structure. The torsional motion and axial motion are naturally coupled, resulting in a smooth and continuous motion trajectory without mechanical impact. By adjusting the arrangement angle, pitch, and excitation intensity of the second component group, the ratio between axial displacement and torsion angle can be independently controlled, enabling precise design of the motion mode. This collaborative mechanism originates from structural design rather than complex control, allowing the composite to spontaneously generate the expected combined motion at the moment of excitation, resulting in fast response and high reliability.

[0306] For example, taking a biomimetic muscle composite made of SBAS material, the first component group consists of 20 axially parallel components with a diameter of 0.3 mm and a modulus of 1.2 MPa, forming a core bundle with a diameter of 2 mm. The second component group consists of 40 components with a diameter of 0.2 mm and a modulus of 0.4 MPa, divided into left and right groups of 20 each, which are cross-wound around the outer periphery of the core bundle with helical angles of 45° and -45° respectively, with 2 layers of winding and a pitch of 3 mm. When only the first component group is excited, the composite exhibits an 8% axial contraction, shortening its length from 100 mm to 92 mm. When only the second component group is excited, the composite exhibits a bending deformation of approximately 15°. When both groups of components are excited simultaneously, the bending driving force of the second component group generates a circumferential component through the helical path, which works synergistically with the axial contraction of the first component group, causing the composite to exhibit a 30° torsion around its axis while undergoing a 5% axial contraction. High-speed imaging and mechanical testing systems recorded and showed that torsional motion and axial contraction occurred synchronously without time delay, and the torsional angle was linearly related to the excitation voltage, which could be continuously adjusted within the range of 100-500V. With one end of the composite fixed and the other end connected to a pointer, it was clearly observed that the pointer moved axially and rotated around the axis simultaneously, visually verifying the synergistic effect of contraction and bending in generating torsion.

[0307] In biomimetic robotic wrist joint actuation applications, this contraction-flexion-torsion coordinated mechanism perfectly simulates the complex movements of the human wrist joint. During implementation, the composite is arranged along the forearm axis, with its proximal end fixed to the radial skeleton and its distal end connected to the hand skeleton, simulating the anatomical structure of the radiocarpal joint. When a "pronation" action (palm-downward rotation) is required, the first component group and a second component group in a specific direction (such as a right-handed coiling group) are simultaneously stimulated, causing the composite to slightly contract axially while simultaneously twisting clockwise, thus rotating the hand. When a "supination" action (palm-upward rotation) is required, the second component group in another direction (left-handed coiling group) is stimulated, producing a counter-clockwise twist. When wrist flexion is required accompanied by slight rotation, the stimulation ratio of the two second component groups is adjusted to combine torsion and flexion. In grasping operations, when the robotic hand needs to adjust the object's posture after grasping it, this coordinated mechanism allows the wrist joint to rotate the object while maintaining gripping force, achieving continuous "grasping and rotation" movements. In actual testing, the bionic wrist joint prototype can achieve an axial contraction range of 0-8mm (corresponding to wrist flexion and extension of 0-60°) and a twisting range of 0-45°. Both can be adjusted independently or in tandem to complete complex wrist movements such as pronation, supination, ulnar deviation, and radial deviation. The movement trajectory is highly similar to the movement of the human wrist joint, significantly improving the flexibility and naturalness of robot operation.

[0308] In some embodiments, the first component group and the second component group respond to electric field excitations with different start times, different durations, or different frequencies, causing the first deformation and the second deformation to form a preset phase relationship, sequence relationship, or frequency combination in time, thereby enabling the flexible biomimetic muscle complex to generate motion output, including at least spiral motion, oscillating motion, or wave motion. The strip-shaped components constituting the first component group and / or the second component group are configured to be independently addressable and driven, and the motion output is achieved by applying differentiated driving electric fields to different strip-shaped components or different component groups.

[0309] In a specific implementation, electrical signals with specific timing parameters are applied to the first and second component groups via independently addressed electrode lines. For example, the start time of the two sets of excitation signals may be delayed (e.g., 50ms), the duration may differ (e.g., the first group lasts 0.5s, the second group lasts 1.0s), or the frequency may differ (e.g., the first group lasts 10Hz, the second group lasts 5Hz). The selection of these timing parameters is based on the dynamic characteristics of the target motion mode: when spiral motion is required, the first and second deformations can be made to have a 90° phase difference, causing axial contraction and bending torsion to overlap in time; when oscillating motion is required, the two sets of deformations can occur alternately, forming a periodic reciprocating motion; when wave motion is required, multiple segments of the composite body can be sequentially excited along the length direction, forming a traveling wave effect. This time-coordination mechanism introduces temporal dimension control on the basis of spatial coordination, enabling the composite to generate richer and more varied dynamic motion modes. By adjusting the timing parameters of the electrical signal, the same composite can quickly switch between different motion modes without changing the hardware structure. Time coordination enables the composite to simulate the temporal recruitment mechanism of biological muscles (such as peristalsis and rhythmic movement), significantly improving the biomimicry realism. This mechanism provides a simple and effective control means for the generation of complex movements (such as spiral propulsion and wave-like swimming), avoiding the complex coordination requirements of multiple independent actuators.

[0310] In one embodiment, the first and second component groups can generate oscillating motion based on different start times. Specifically, when the electric field excitation start times of the first and second component groups are different, they can be arranged in a sequential relationship in time, thereby generating oscillating motion. Taking a biomimetic finger composite made of SBAS material as an example, the first component group (axial contraction) and the second component group (bending) are controlled separately by independent electrodes. The excitation parameters are set as follows: the first component group is applied with a voltage of 500V for 1 second; the second component group is applied with the same voltage, but with a 0.2s delay in start time. When the excitation begins, the first component group first generates axial contraction, shortening the finger as a whole; after 0.2s, the second component group begins to generate bending deformation. At this time, the finger is already in a contracted state, and the bending deformation superimposed on the contraction causes the finger to swing to one side; after the excitation ends, the two components relax synchronously, and the finger returns to its original position. By adjusting the delay time (e.g., 0.1s, 0.3s, 0.5s), the amplitude and speed of the swing can be controlled. High-speed video recordings showed that the finger swing amplitude reached 25° at a delay of 0.2s, and decreased to 15° at a delay of 0.5s, proving the regulatory effect of the initial time difference on the swing motion.

[0311] In the simulation of a "tapping" motion in a biomimetic robot finger, this oscillation mechanism based on an initial time difference can produce a movement similar to a human finger tapping a table. In implementation, a composite element serves as the driving unit for the index finger, with its proximal end fixed to the hand skeleton and its distal end connected to the fingertip. When a tapping motion is required, the control system first excites the first component group to cause axial contraction of the finger (simulating flexion), and after a 50ms delay, excites the second component group to cause lateral oscillation of the finger (simulating tapping direction adjustment). The "crispness" of the tap can be controlled by adjusting the delay time; a shorter delay (e.g., 20ms) produces a rapid tap, while a longer delay (e.g., 100ms) produces a gentle stroking motion. In applications with performance robots, combining this mechanism with multi-finger collaborative control allows each finger to oscillate with different initial time differences, simulating the timing and force variations of different fingers touching the keys in piano playing, significantly improving the naturalness and expressiveness of the performance.

[0312] In one embodiment, the first and second component groups can generate wave motion based on different durations. Specifically, wave motion can be generated when the electric field excitation durations of the first and second component groups are different and they are segmented along the length of the composite. Taking a biomimetic worm robot as an example, a composite with a length of 200 mm is divided into 5 independent control segments (40 mm each) along its length. Each segment contains a first component group (axial contraction) and a second component group (bending). The excitation sequence is set as follows: starting from the first segment, the first component group is excited for 0.5 s, and the second component group is excited for 0.3 s. The two excitations start simultaneously, but the second group ends first. After the first segment ends, the second segment starts with the same excitation mode after a delay of 0.1 s, and so on until the fifth segment. This excitation sequence causes each segment of the composite to bend briefly while contracting axially, and the bending ends before the contraction ends, forming a local deformation sequence of "contraction, bending, relaxation". The deformation of each segment is transmitted sequentially along the length, and the whole presents a wave-like peristalsis. High-speed camera footage shows that this excitation mode generates traveling waves with a wavelength of 40 mm and a wave speed of about 80 mm / s on the surface of the composite, propelling the biomimetic worm forward on a horizontal plane.

[0313] In the autonomous propulsion application of gastrointestinal endoscopic robots, this wave motion mechanism based on different durations can simulate the propulsion mode of intestinal peristalsis. In implementation, the composite is fabricated into a slender tubular structure with a diameter of 10 mm and a length of 150 mm, divided into 8 independently controllable segments along its length. When forward propulsion is required, each segment is excited sequentially according to the above timing sequence: the ratio of contraction to bending duration in each segment is 5:3, with a 50 ms delay between segments. This wave-like deformation causes periodic bulges on the surface of the composite, generating friction with the intestinal wall and propelling the robot forward. By adjusting the excitation duration and the inter-segment delay, the propulsion speed and direction can be controlled. For example, shortening the inter-segment delay increases the wave speed, raising the propulsion speed from 2 mm / s to 5 mm / s; changing the excitation sequence allows the robot to move in the opposite direction. In in vitro intestinal model testing, this wave motion mechanism successfully drove the robot through a tortuous intestinal path, effectively improving the average propulsion efficiency compared to traditional peristalsis methods, and reducing the peak pressure on the intestinal wall, significantly improving safety and passage.

[0314] In one embodiment, the first and second component groups can generate helical motion based on different frequencies. Specifically, when the electric field excitation frequencies of the first and second component groups are different, they can form a periodic phase change over time, thereby generating helical motion. Taking a bionic wrist joint complex as an example, a 10Hz sinusoidal voltage is applied to the first component group (axial contraction), and a 5Hz sinusoidal voltage is applied to the second component group (bending and torsion), with an initial phase difference of 0°. Due to the different frequencies, the phase difference between the two deformations changes periodically with time. Specifically, when the phase difference is 0°, axial contraction and bending and torsion are superimposed in the same direction, producing the maximum torsional amplitude; when the phase difference becomes 180°, they cancel each other out in opposite directions, resulting in the minimum torsional amplitude; this cycle repeats, causing the torsional angle of the complex to change periodically with time, while the axial length also pulsates at a frequency of 10Hz, resulting in an overall spatial motion along a helical path. Mechanical tests show that this frequency combination causes the end of the complex to trace a helical trajectory with a diameter of 5mm and a pitch of 2mm within 10 seconds. By adjusting the frequency ratio (such as 2:1, 3:1, 5:2), spiral motions with different densities can be generated.

[0315] In the guidewire drive of vascular interventional robots, this helical motion mechanism based on different frequencies can simulate the rotational advancement of the guidewire during vascular interventional surgery. In implementation, the composite is fabricated into an ultra-fine guidewire with a diameter of 1 mm and a length of 300 mm. Its first component group provides axial contraction (10 Hz excitation), and the second component group provides bending and torsion (5 Hz excitation). When traversing a curved section of the vessel, this frequency combination is activated to generate a helical motion at the tip of the guidewire. Its rotational effect guides the guidewire along the vessel centerline, reducing lateral pressure on the vessel wall. When precise control of the guidewire direction is required, switching to the same frequency excitation stops the helical motion, retaining only axial contraction for linear advancement. In silicone vessel model testing, this helical motion mechanism improved the success rate of the guidewire traversing a 90° bend compared to traditional linear advancement methods, and reduced the peak contact force with the vessel wall. Furthermore, by adjusting the frequency ratio, it can adapt to different vessel diameters. Specifically, large-diameter vessels use a smaller frequency ratio (e.g., 2:1) to generate a large-pitch helix, while small-diameter vessels use a larger frequency ratio (e.g., 5:1) to generate a denser helix, achieving personalized adaptation.

[0316] In some embodiments, at least a portion of the multiple strip members are configured to be independently addressable and driven, thereby achieving motion output by applying differentiated driving electric fields to different strip members or groups of members; or the strip members constituting the first group of members and / or the second group of members are configured to be independently addressable and driven, thereby achieving motion output by applying differentiated driving electric fields to different strip members or groups of members.

[0317] In a specific implementation, by setting independent electrode circuits for each or each group of strip-shaped components, they can receive different electric field excitations, thereby achieving fine-grained control of the driving units within the composite. Specifically, independent addressing driving can be implemented at two granularities: one is component-level independent addressing, where each strip-shaped component is connected to an independent electrode channel, and the controller can apply electrical signals with different voltage amplitudes, waveforms, frequencies, or phases to each component; the other is component group-level independent addressing, where the strip-shaped components constituting the first or second component group are controlled by independent electrode buses, with synchronous excitation within the group but independent adjustment of excitation parameters between groups. Through this differentiated driving electric field configuration, driving units at different locations within the composite can work collaboratively according to preset spatial and temporal patterns. For example, in a bionic finger, component groups of different phalanges can be independently excited to produce differences in flexion timing, simulating the progressive bending of the finger; or in a multi-degree-of-freedom joint, by independently adjusting the excitation intensity of component groups in different orientations, the joint can be deflected in any direction in three-dimensional space. This independently addressable driving method provides extremely high programmability for the motion modes of flexible biomimetic muscle complexes, enabling the same complex to achieve multiple different motion outputs by changing the electrical signal configuration without altering the hardware structure. Component-level fine control can compensate for potential performance differences during fabrication, improving the consistency and accuracy of the overall motion of the complex. This mechanism provides the hardware foundation for the subsequent introduction of time coordination (such as different start times and frequencies), enabling the complex to generate complex dynamic motions such as oscillations, waves, and spirals. The independent addressing capability allows the complex to adapt to different loads and operating conditions, such as enhancing the driving force in specific areas when encountering resistance, thereby improving the robustness and adaptability of the driving system.

[0318] In some embodiments, the motion output may include at least one of longitudinal (axial) contraction, multi-directional bending, torsion, or any combination thereof.

[0319] In a specific implementation, within the flexible biomimetic muscle complex, various motion outputs and their combinations can be achieved through different excitation states and spatial structural configurations of the first and second component groups. Longitudinal (axial) contraction is primarily achieved by individually exciting the first component group. Correspondingly, when the high-modulus strip-shaped component in the first component group undergoes dominant axial contraction under electric field excitation, the entire complex shortens along the axis; if the first component group uses an elongation-type electro-actuated material, axial elongation can be achieved. Multi-directional bending motion is achieved by individually or in combination exciting the second component group. Correspondingly, when the low-modulus component in the second component group undergoes bending deformation, the complex can bend in different directions depending on its spatial distribution within the complex (e.g., offset to one side, uniformly distributed circumferentially, or arranged at a specific angle); for example, concentrating the second component group on one side of the complex allows for unidirectional bending to the opposite side upon excitation; if the second component group is arranged in uniformly circumferential sections, multi-directional bending control can be achieved by selectively exciting different areas. The torsional motion is achieved through the spatial coordination of the first and second component groups. Correspondingly, when the second component group wraps around the periphery of the first component group in a spiral or cross-spiral manner, its bending driving force generates a circumferential component, which, together with the axial contraction of the first component group, drives the composite to rotate around its own axis. The direction and amplitude of the torsion can be controlled by adjusting the spiral angle and the winding direction. For combined motion, it is achieved by simultaneously or sequentially exciting the two groups of components. For example, simultaneously exciting the first and second component groups can cause the composite to bend or twist while contracting axially, realizing a combined motion of contraction and bending or a combined motion of contraction and torsion. If a time coordination mechanism is further introduced (such as different start times or different frequencies), more complex motion sequences can be generated, such as oscillating motion with contraction followed by bending, or spiral motion with axial pulsation and torsion coupling.

[0320] In some embodiments, the flexible biomimetic muscle complex is configured to achieve independent and continuous control of the motion output of the flexible biomimetic muscle complex by adjusting the electrical signal parameters applied to the first component group and / or the second component group.

[0321] In a specific implementation, an independently addressable drive architecture is used to apply independently adjustable electrical signals to the first and second component groups. For example, adjusting the voltage amplitude can change the deformation amplitude (a larger amplitude means stronger contraction or a larger bending angle), adjusting the signal frequency can change the response speed and dynamic characteristics (low frequencies produce continuous contraction, while high frequencies produce vibration or creep), adjusting the pulse width or duty cycle can control the deformation holding time, and adjusting the phase difference can achieve time-adjusted or time-lag coordination between the two sets of deformations. By continuously adjusting these parameters, the motion output of the composite can be steplessly adjusted in dimensions such as axial contraction, bending angle, and torsional amplitude, and smoothly transition between different motion modes. This system enables precise control of motion output, allowing the composite to adjust its posture and output force in real time based on external commands or sensor feedback, meeting the demands of high-precision operation. "Independent" control allows for separate optimization of the contributions of the first and second component groups; for example, the first group's excitation can be enhanced when strong contraction is needed, while the second group's excitation can be enhanced when fine bending is required, without interference between the two. "Continuous" adjustment avoids the motion shock and unevenness caused by step-by-step control, making the composite's motion trajectory more natural and smooth, closely resembling the gradual contraction characteristics of biological muscles. This control mechanism provides an interface for the application of adaptive control and machine learning algorithms, enabling the composite to autonomously adapt to different loads and environmental conditions by optimizing electrical signal parameters online.

[0322] In some embodiments, the plurality of component groups further includes a third component group, the elastic modulus of the third component group being greater than that of the second component group and less than that of the first component group; the first component group, the third component group and the second component group have a preset second spatial structure relationship, the second spatial structure relationship including at least one of the following: second attachment type, second covering type, second embedded type, and second spaced type.

[0323] In a specific embodiment, the flexible biomimetic muscle composite may further include a third component group, whose elastic modulus is configured to be greater than that of the second component group and less than that of the first component group, thereby introducing an intermediate modulus transition layer between the high and low modulus groups. This third component group is achieved by controlling the preparation process (e.g., using a moderate stretching ratio of 3:1, moderate filler content, or appropriate crosslinking density) to obtain strip-shaped components (e.g., 0.7-0.8 MPa) with modulus values ​​between the first component group (e.g., 1.2 MPa) and the second component group (e.g., 0.4 MPa), and then independently bundling them into groups.

[0324] In terms of spatial arrangement, the third component group can be coupled with the first and second component groups according to the preset second spatial structural relationship. By introducing the intermediate modulus component group, the modulus distribution of the composite is expanded from binary discrete to ternary gradient, providing more degrees of freedom for the fine-tuning of motion modes. Secondly, the third component group can act as a "motion smoothing adjustment layer," providing a mechanical transition between contraction drive and bending drive, making the motion of the composite more smooth and continuous when simultaneously excited, avoiding motion shocks caused by sudden changes in modulus. By independently controlling the excitation intensity of the third component group, fine-tuning of the intermediate state of the composite motion can be achieved, such as adjusting the ratio of axial contraction to torsion in a contraction and torsional composite motion.

[0325] Second attachment type: The third component group is attached to the first component group, and the second component group is attached to the third component group.

[0326] Figure 15 This is a schematic diagram of a second attachment structure provided in one embodiment of this application.

[0327] Among them, reference Figure 15 As shown, in one embodiment, the third component group C3 is attached to the surface of the first component group C1, and the second component group C2 is then attached to the surface of the third component group C3, forming a series structure with layers attached from the inside out. Specifically, the implementation is as follows: First, a high-modulus strip-shaped component of the first component group is prepared as the core layer. An uncured intermediate-modulus material slurry or interfacial coupling agent is coated on its surface. The strip-shaped component of the third component group is then longitudinally attached to it, and an integrated interface is formed through co-curing. After the third component group has cured, a low-modulus material slurry is coated on its surface, and the strip-shaped component of the second component group is attached to the outermost layer. Co-curing is then performed again to ensure a continuous bond between all interfaces. Unlike the first attachment method (where the second component group is directly attached to the first component group), the second attachment method achieves indirect attachment through the third component group as an intermediate transition layer, resulting in a radial "high, medium, low" modulus gradient distribution among the three component groups. The intermediate modulus transition layer provided by this structure alleviates the abrupt change in stiffness between high-modulus and low-modulus materials, making the interfacial stress distribution more uniform and significantly improving the composite's resistance to interfacial delamination under dynamic driving. The three-layer attachment structure provides more adjustable interfaces. By independently controlling the material properties of the third component group (such as thickness and modulus value), the force transmission efficiency and coupling strength between the first and second component groups can be finely adjusted. This structure enables the three components to be tightly integrated in space, providing a structural basis for subsequent realization of more complex gradient functional distributions (such as continuous modulus changes). Compared with the first attachment type, the second attachment type can realize more diverse driving force combinations. For example, the first component group dominates contraction, the second component group dominates bending, and the third component group, as an intermediate adjustment layer, can be independently excited to smooth the movement of both.

[0328] For example, taking the preparation of a biomimetic muscle composite using SBAS material as an example, the first component group uses 20 high-strength (5:1) strip components with a diameter of 0.3 mm and a modulus of 1.2 MPa bundled in parallel as the core layer; the third component group uses 10 medium-strength (3:1) strip components with a diameter of 0.2 mm and a modulus of 0.7 MPa, after coating their surfaces with a dilute SBAS solution, and then attaching them longitudinally to the outer periphery of the first component group, and pre-curing them at 100°C for 10 minutes; the second component group uses 20 low-strength (2:1) strip components with a diameter of 0.2 mm and a modulus of 0.4 MPa, after coating the surface of the third component group with a coupling agent, and then wrapping them in a 45° spiral manner to attach them to the outermost layer, and then co-curing them at 120°C for 20 minutes, so that the three-layer interface forms a continuous chemical cross-link. Cross-sectional observation of the fabricated composite showed good interface fusion between the three layers with no obvious boundaries. Mechanical tests indicated that its interfacial shear strength reached 6.8 MPa, which is improved compared to the directly attached structure without an intermediate layer (4.2 MPa). In electric field excitation tests, 8% axial contraction was generated when only the first component group was excited; 25° bending was generated when only the second component group was excited; when the first and third component groups were excited simultaneously, the axial contraction and intermediate layer contraction synergistically increased the contraction rate to 10%; when all three groups of components were excited simultaneously, the axial contraction, intermediate layer adjustment, and bending deformation synergistically produced a smooth composite contraction and bending motion with no impact vibration during the bending process, which significantly improved the smoothness of motion compared to the two directly attached structures.

[0329] In the application scenario of rotating and flexing joint actuation in the forearm of a biomimetic robot, the second attached structure can simulate the coordinated movement of the radius and ulna in the human forearm. During implementation, the first component group (high modulus) corresponds to the main actuation of the radius, responsible for axial support and strong contraction of the forearm; the third component group (medium modulus) corresponds to the interosseous membrane region, serving as a mechanical transition and smoothing layer; and the second component group (low modulus) corresponds to the flexion actuation of the ulnar side, responsible for bending deformation. When a combined "pronation and elbow flexion" movement is required, both the first and second component groups are simultaneously stimulated, causing the forearm to flex while contracting axially, while the third component group remains passive, acting only as a transition layer. When fine adjustment of the ratio of flexion angle to contraction force is needed, the third component group is additionally stimulated. This enhanced stimulation increases the stiffness of the intermediate layer, allowing more contraction force to be transmitted to the bending end, achieving fine-tuning of the flexion angle. In grasping operations, when the robotic arm grasps fragile items, the independent adjustment of the third component group ensures that the composite maintains the necessary gripping force while possessing appropriate flexibility, avoiding impact damage. This design makes the movement of the forearm joint smoother and more natural, with a movement trajectory that is more than 90% similar to that of the human forearm. It can also achieve a continuous transition from strong gripping to fine operation through the independent control of three sets of components.

[0330] Second type of encapsulation: The third component group encapsulates the first component group, and the second component group encapsulates the third component group.

[0331] Figure 16 This is a schematic diagram of a second encapsulated structure provided in one embodiment of this application.

[0332] Among them, reference Figure 16 As shown, in one embodiment, the third component group C3 is spirally or woven around the outer periphery of the first component group C1, and the second component group C2 is then spirally wrapped around the outer periphery of the third component group C3 in the same or different ways, forming a concentric layered structure with layering from the inside out. Specifically, the implementation is as follows: First, the high-modulus first component group strip is used as the core layer. A continuous winding process is used to spirally wrap the medium-modulus third component group around its surface at a preset angle (e.g., 30°-60°) and pitch, forming the first coating layer. Then, a coupling agent or a slurry of the same material is coated onto the surface of the first coating layer. Next, the low-modulus second component group is spirally wrapped around the surface of the third component group at the same or opposite angle, forming the second coating layer. Finally, co-curing allows the molecular chains at the interfaces of each layer to diffuse and fuse, forming an integrated multilayer composite structure. Unlike the first coating type (where the second component group directly wraps the first component group), the second coating type achieves a gradient transition of modulus from the inside to the outside (high, medium, low) through the intermediate layer of the third component group. The intermediate modulus cladding layer provided by this structure alleviates the abrupt stiffness change between the high-modulus core layer and the low-modulus outer layer, making the radial stress distribution more uniform and significantly improving the composite's resistance to interfacial delamination during bending and torsion. The multi-layer cladding structure provides more adjustable geometric parameters. By independently setting the cladding angle, pitch, and thickness of each layer, independent adjustment of axial contraction, torsional stiffness, and bending flexibility can be achieved. The third component group, as an intermediate layer, can be independently excited, acting as a "stiffness regulating valve." That is, exciting the intermediate layer can change its modulus, thereby adjusting the efficiency of the transmission of the driving force from the outer layer to the inner layer. Compared with the first cladding type, this structure can achieve more complex composite motions, such as the outer layer driving torsion, the inner layer driving contraction, and the intermediate layer adjusting the coupling strength between the two, making the motion mode of the composite more diverse and adjustable.

[0333] For example, a second coated structure is prepared using a liquid crystal elastomer and carbon nanotube composite material. The first component group uses a thermotropic nematic liquid crystal elastomer (modulus 1.5 MPa), which is electrospun into fibers with a diameter of 0.2 mm and bundled in parallel as the core layer. The third component group uses an azobenzene-doped photoresponsive liquid crystal elastomer (modulus 0.8 MPa), which is prepared into fibers with a diameter of 0.1 mm and wound around the surface of the core layer with a 45° helix angle and a pitch of 1 mm to form the first coating layer. The azobenzene is then photocrosslinked and oriented by ultraviolet light irradiation. The second component group uses a conductive polypyrrole / liquid crystal elastomer composite material (modulus 0.3 MPa), which is prepared into fibers with a diameter of 0.1 mm and wound around the surface of the third component group with a -45° helix angle and a pitch of 0.8 mm to form the second coating layer. Finally, the three layers are fused together by hot pressing at 120°C for 30 minutes. Multi-mode drive tests were conducted on the composite: when only the first component group was excited (heated to 120°C), 12% axial contraction was generated; when only the second component group was excited (3V voltage was applied), 30° bending was generated accompanied by 10° torsion; when the first and second component groups were excited simultaneously, axial contraction and bending torsion were combined; when the third component group was additionally excited (excited by 365nm ultraviolet light), the modulus of its intermediate layer decreased to 0.5MPa due to photo-induced isomerization, which improved the efficiency of the transmission of the driving force from the outer layer to the inner layer and increased the torsion angle, demonstrating the regulating effect of the intermediate layer on the motion output.

[0334] In the guidewire drive system of a vascular interventional robot, the second encapsulated structure can simulate the adaptive expansion and torsion behavior of blood vessels under changes in blood pressure. In implementation, the composite is fabricated into a guidewire with a diameter of 0.8 mm and a length of 300 mm. The first component group (high modulus) uses a thermoresponsive liquid crystal elastomer to provide axial rigidity support, the third component group (medium modulus) uses a photoresponsive material as an intermediate adjustment layer, and the second component group (low modulus) uses an electroactive conductive polymer to provide bending and torsion drive. When the guidewire needs to pass through a curved section of the blood vessel, the second component group is excited to generate a helical torsion, guiding the guidewire to rotate and advance along the centerline of the blood vessel. When precise guidance is required at a blood vessel bifurcation point, both the first and second component groups are excited to bend and torsion the guidewire head, adjusting the direction of advance. When it is necessary to adapt to changes in blood vessel diameter, the modulus of the third component group is adjusted by ultraviolet light—reducing the intermediate layer modulus in stenotic sections to make the guidewire more compliant, and increasing the intermediate layer modulus in dilated sections to enhance guidewire stiffness. In silicone vascular model testing, this multi-layered guidewire demonstrated a high success rate in navigating complex bending paths and could adapt to dynamic vascular diameter changes of 2-8 mm through adjustment of the intermediate layer, improving its adaptability compared to traditional single-layer driven guidewires. Furthermore, the photoresponsive characteristics of the third component group allow for remote stiffness control of the guidewire within the body via optical fiber, eliminating the need for additional electrical connections. This makes it particularly suitable for interventional medical scenarios with high electromagnetic compatibility requirements.

[0335] Second Embedding: The third component group is embedded inside the first component group, and the second component group is embedded inside the third component group.

[0336] Figure 17 This is a schematic diagram of a second embedded structure provided in one embodiment of this application.

[0337] Among them, reference Figure 17 As shown, in one embodiment, the third component group C3 is embedded inside the first component group C1, and the second component group C2 is then embedded inside the third component group C3, forming a nested structure with layers embedded from the outside in. Specifically, the implementation is as follows: First, a first component group matrix with longitudinal through holes or grooves is prepared (e.g., a high-modulus rod with channels is formed using a co-extrusion process). Then, the material of the third component group is injected into the channels in a molten state or prepolymer form, and after solidification, a first layer of embedding is formed. Next, finer micropores are processed longitudinally inside the third component group, and the material of the second component group is injected and solidified to form a second layer of embedding. Alternatively, a multi-step co-extrusion process can be used to achieve concentric nesting of the three layers of material in a single molding process. Unlike the first embedded structure (where the second component group is directly embedded in the first component group), the second embedded structure uses the third component group as an intermediate transition layer, achieving a gradient distribution of modulus from the outside in (high modulus of the first group → medium modulus of the third group → low modulus of the second group), and the three components are completely coaxial in the radial direction with continuous interfaces. This structure creates omnidirectional force coupling in three-dimensional space. The outer high-modulus matrix provides structural support and the main drive for axial contraction, the middle layer regulates stress transmission, and the inner low-modulus component undergoes fine bending or torsion when constrained by the outer layer. The interaction of these three elements makes the force transmission path shorter and more efficient. The embedded structure completely protects the low-modulus component inside the high-modulus matrix, significantly improving the composite's tensile strength, fatigue resistance, and environmental resistance. By designing the spatial distribution of the channels (such as eccentric holes or multi-hole arrays), the point and direction of action of the inner component on the outer matrix can be precisely controlled, realizing composite motion in three-dimensional space. This structure allows the composite to integrate multiple functional units while maintaining a small outer diameter, resulting in extremely high space utilization, making it particularly suitable for miniaturized applications.

[0338] A second embedded structure was fabricated using an ionomer-metal composite (IPMC) and a carbon fiber reinforced epoxy resin composite as an example. The first component group was made of carbon fiber reinforced epoxy resin (modulus 15 GPa), pultruded into a 2 mm diameter rod, with three 0.5 mm diameter eccentric holes (circumferentially distributed at 120°) machined axially inside. The third component group was made of PVDF piezoelectric polymer (modulus 1.2 GPa) doped with barium titanate nanoparticles, injected into the eccentric holes in a molten state, polarized under a high-voltage electric field to orient the molecular chains axially, and cooled and solidified to form the first embedded layer. Subsequently, micropores with a diameter of 0.2 mm were machined along the same axis inside the third component group, and IPMC slurry (a composite of perfluorosulfonic acid resin and platinum nanoparticles) was injected into the micropores. Electrodes were prepared by chemical reduction to form the second embedded layer. The resulting composite had a total diameter of 2.2 mm and contained three independent IPMC driving units. Driven testing showed that the composite remained rigid when the first component group (the carbon fiber composite material itself is not actively driven, but serves as a structural layer) was excited. When any IPMC core was excited (3V voltage was applied), the core bent. Due to its eccentric embedding, the bending force exerted local stress on the outer matrix, causing the composite to bend to the opposite side. Simultaneously exciting the three IPMC cores and controlling the voltage amplitude differences allowed the composite to bend in any direction in three-dimensional space, with a maximum bending angle of ±45°. In addition, the PVDF layer of the third component group generated piezoelectric signals when the composite deformed, which could monitor the bending state in real time and achieve a self-sensing function.

[0339] In the application of multi-degree-of-freedom grasping instruments in minimally invasive surgical robots, the second embedded structure enables high-precision, multi-angle manipulation of tissues. In implementation, the composite is fabricated into a surgical forceps drive rod with a diameter of 2mm and a length of 50mm. Its first component group (carbon fiber reinforced composite material) provides basic rigidity to prevent rod buckling, the third component group (PVDF piezoelectric layer) serves as an intermediate transition and self-sensing layer, and the second component group (three IPMC inner cores) acts as the active drive unit, controlling the opening, closing, deflection, and rotation of the forceps head. When tissue needs to be grasped, the three IPMC inner cores are simultaneously stimulated to produce symmetrical bending, driving the forceps head to close. When the grasping angle needs to be adjusted, the three inner cores are stimulated differently; for example, increasing the stimulation of one inner core causes the forceps head to deflect to that side. When fine tissue dissection is required, the contact force between the forceps head and the tissue is monitored in real time through the PVDF layer, and the stimulation intensity of the IPMCs is adjusted accordingly to achieve force-controlled operation. In simulated surgical tests, the instrument successfully performed delicate operations such as dissecting blood vessels with a diameter of 1 mm and grasping and rotating soft tissues. Its multi-degree-of-freedom control capability effectively improves surgical precision, and the PVDF self-sensing layer allows the operator to perceive tissue hardness and contact force in real time, significantly reducing the risk of misoperation. In addition, the fully embedded structure makes the instrument surface smooth and without protrusions, facilitating cleaning and disinfection, and meeting the hygiene requirements for medical devices.

[0340] The second type of alternating arrangement: the first component group, the third component group, and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction.

[0341] Figure 18 This is a schematic diagram of a second spaced structure provided in one embodiment of this application.

[0342] In one embodiment, after the introduction of the third component group, the first component group, the third component group, and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction, forming a periodic multi-layered composite structure. For example... Figure 18 As shown, composite G2 is obtained by stacking components in the following order from top to bottom: "First component group C1, Second component group C2, Third component group C3, Second component group C2, First component group C1". Specifically, it is implemented using a layered assembly or multi-channel co-extrusion process, where the high-modulus first component group, the medium-modulus third component group, and the low-modulus second component group are stacked alternately in the thickness direction, one layer at a time, in a preset order. For example, they can be arranged in a symmetrical or asymmetrical sequence such as "First, Third, Second, Third, First" or "First, Second, Third, Second, First". The number of stacked layers can be designed as an odd number, such as 5, 7, or 9 layers, to ensure symmetry on both sides. The interfacial molecular chains diffuse and fuse between the layers through hot pressing, forming an integrated multi-layered alternating composite structure. Unlike the first interleaved type (only two alternating groups), the second interleaved type introduces an intermediate modulus transition layer, causing the modulus to exhibit a multi-level gradient change in the thickness direction, such as "high, medium, low, medium, high" or "high, low, medium, low, high". The periodic alternating arrangement of this structure ensures highly uniform mechanical properties of the composite in the thickness direction, resulting in a smoother stress distribution and avoiding localized stress concentration caused by abrupt changes in modulus. The introduction of an intermediate modulus layer provides a mechanical buffer between high and low moduli, significantly improving the composite's resistance to delamination during dynamic bending. By adjusting the alternation sequence and layer ratio, the overall bending stiffness, neutral layer position, and deformation symmetry of the composite can be independently controlled. For example, a three-layer structure of "high, low, high" can produce unidirectional bending, while a five-layer structure of "high, medium, low, medium, high" can achieve bidirectional symmetrical bending. This structure provides a feasible path for the subsequent construction of more complex gradient functional materials (such as continuous gradient distributions) through discretization approximation.

[0343] A second spacer structure was prepared using a piezoelectric ceramic fiber composite material and a conductive polymer composite material as an example. The first component group used lead zirconate titanate piezoelectric ceramic fiber (modulus 60 GPa), with a diameter of 0.1 mm and a length of 50 mm, coated with silver electrodes. The third component group used barium titanate-doped PVDF piezoelectric polymer fiber (modulus 2 GPa), with a diameter of 0.15 mm. The second component group used polypyrrole / polyurethane conductive polymer fiber (modulus 0.1 GPa), with a diameter of 0.2 mm. A laminated assembly process was used, with 5 layers arranged alternately in the thickness direction in the order of "first, third, second, third, first", with 10 fibers laid in each layer. Epoxy resin adhesive was applied between the layers, and the layers were hot-pressed at 80℃ and 0.3 MPa for 1 hour to cure, resulting in a layered composite strip (hereinafter referred to as the composite strip) with a thickness of approximately 1.2 mm and a width of 5 mm. The composite strip was subjected to a driving test: when a high-voltage electric field (1kV / mm) was applied to the first component group, the piezoelectric ceramic underwent axial contraction, causing the composite strip to shorten overall; when a 3V voltage was applied to the second component group, the conductive polymer bent, and due to its intermediate layer position, the bending force was transmitted to the first component group through the third component group, causing the composite strip to produce a smooth S-shaped bend; when all three layers were excited simultaneously, the contraction of the piezoelectric ceramic and the bending of the conductive polymer worked synergistically, causing the composite strip to produce a wavy deformation. Furthermore, the PVDF layer of the third component group generated a piezoelectric signal during the deformation of the composite strip, allowing for real-time monitoring of the bending curvature and achieving a self-sensing function.

[0344] In the application of biomimetic fish fin wave propulsion, the second-interval structure can simulate the alternating stiffness distribution and wave motion of the fin rays. In implementation, the composite is fabricated as a fin array with a thickness of 1 mm, a width of 10 mm, and a length of 200 mm. Each fin ray employs a five-layer alternating structure of "high, medium, low, medium, high". Multiple fin rays are arranged in parallel to form a fish fin. The first component group (piezoelectric ceramic) of each fin ray is connected to a high-voltage drive circuit, the second component group (conductive polymer) is connected to a low-voltage drive circuit, and the third component group (PVDF) serves as a self-sensing layer. When a propulsive wave is needed, the second component groups of each fin ray are sequentially excited, specifically from back to front with a 50 ms delay, causing the fin ray to generate a sequentially transmitted bending wave, propelling the water flow and generating propulsion. Simultaneously, the overall stiffness of the fin ray can be adjusted by the synchronous contraction of the first component group; increased stiffness during contraction makes the wave steeper, improving propulsion efficiency. When rapid turning is required, the second component groups of the fin ray on both sides are differentially excited, causing the fish fin to bend asymmetrically, achieving a turning torque. In pool testing, the biomimetic fish fin prototype achieved a swimming speed of 0.5 m / s and a turning radius of only 0.3 m. Furthermore, the PVDF self-sensing layer allows for real-time monitoring of fin deformation, and closed-loop control improves propulsion efficiency. In addition, the alternating interval structure ensures structural stability of the fins during repeated undulations, and no interlayer delamination was observed after extensive cyclic testing, meeting the requirements for long-term underwater operations.

[0345] In some embodiments, the elastic moduli of the first component group, the third component group, and the second component group are distributed in a continuous gradient along the radial direction on the cross-section of the composite.

[0346] In a specific implementation, the modulus value exhibits a smooth, monotonically decreasing change from the central axis of the composite to the outer periphery. That is, the modulus is highest in the central region (corresponding to the first component group), medium in the middle region (corresponding to the third component group), and lowest in the outer periphery region (corresponding to the second component group). Furthermore, there is no obvious interface between adjacent regions, and the modulus change curve is continuously differentiable. The specific implementation can be achieved by using a gradient curing process. Specifically, during the composite preparation process, the modulus can be continuously varied by controlling the crosslinking density, filler concentration, or molecular orientation in the radial direction. For example, a centrifugal casting process can be used to inject liquid polymers containing different proportions of high-modulus fillers (such as carbon nanotubes) layer by layer from the center to the periphery. Under the action of the centrifugal force field, the filler concentration naturally forms a gradient distribution in the radial direction. Alternatively, a variable temperature curing process can be used to allow the composite to experience different thermal histories from the center to the periphery. The central region has a high curing temperature and a high crosslinking density, while the peripheral region has a low curing temperature and a low crosslinking density, thereby forming a continuous modulus gradient. Magnetic field-induced gradient assembly can also be used, where a radial gradient magnetic field is applied during the composite curing process, causing the magnetic filler to form a concentration gradient distribution in the radial direction. This structure eliminates the unavoidable interlayer interfaces in discrete layered structures, avoiding interfacial delamination failure and significantly improving the fatigue life of the composite under repeated dynamic driving. The continuous gradient distribution allows stress to be smoothly transmitted in the radial direction without any stress concentration points, making the composite deform more uniformly and naturally when subjected to combined loads such as bending and torsion. This structure is closer to the natural mechanical transition from muscle fibers to connective tissue in biological muscles, resulting in higher biomimicry. Fourth, by controlling the shape of the gradient curve (such as linear gradient, S-shaped gradient, exponential gradient), the macroscopic bending stiffness, torsional stiffness, and motion mode of the composite can be precisely designed, providing a material basis for advanced biomimetic motion.

[0347] In some embodiments, the first component group, the second component group, and the third component group work together through a preset spatial arrangement configuration to jointly simulate the functional division and spatial topological relationship of different types of muscle fibers in biological muscles.

[0348] In one embodiment, a high-modulus first component group, corresponding to fast-twitch muscle fibers (responsible for strong contraction and rapid response), is densely arranged along the functional axis of the complex to form the main driving skeleton. A low-modulus second component group, corresponding to slow-twitch muscle fibers (responsible for fine bending and posture adjustment), is distributed around the first component group in a specific spatial configuration (such as spiral wrapping, eccentric embedding, or intermittent arrangement). A medium-modulus third component group, corresponding to intermediate muscle fibers, serves as a transition layer between the two, spatially manifested as wrapping around the periphery of the first component group, embedding inside the first component group, or alternating with the two. Through force transmission structures (such as co-curing interfaces, interlacing connection points, or shared anchor points), the three components form an integrated mechanical coupling. When excited by an electric field, the three components spontaneously generate different deformations based on their respective intrinsic modulus characteristics. Specifically, the first group dominates axial contraction, the second group dominates bending and torsion, and the third group plays a role in smooth transition and stiffness adjustment. The deformations of the three groups are naturally superimposed in space, realizing a continuous spectrum of output from rapid and powerful motion to fine adjustment. This biomimetic arrangement and configuration, working in synergy, achieves for the first time in artificial muscles a leap from binary functional differentiation to ternary fine control, more realistically simulating the multi-type synergistic mechanisms of fast-twitch, slow-twitch, and intermediate-type muscle fibers in biological muscles. The third component group, acting as a mechanical buffer layer, eliminates the abrupt change in stiffness between high and low modulus, making the motion of the composite more smooth and natural during dynamic driving, approaching the gradual contraction characteristics of biological muscles. By independently controlling the excitation state of the third component group, the overall stiffness and motion mode of the composite can be adjusted in real time, achieving a continuous switch from "rigid burst" to "flexible compliance". The spatial topology design of the three components provides a universal framework for constructing more complex biomimetic structures (such as multipennate muscles and circular muscles), enabling the composite to simulate various muscle types from skeletal muscle to cardiac muscle.

[0349] For example, taking the biomimetic simulation of the human biceps brachii as an example, its anatomical structure includes fast-twitch fibers (dominantly responsible for explosive contraction), slow-twitch fibers (dominantly responsible for endurance flexion), and intermediate fibers (regulating the synergy between the two). Three components are constructed using SBAS material: the first component group (fast-twitch simulation) consists of 30 high-strength fibers with a draw ratio of 5:1 and a modulus of 1.2 MPa, bundled in parallel and arranged along the functional axis, responsible for generating strong axial contraction; the second component group (slow-twitch simulation) consists of 20 low-strength fibers with a draw ratio of 2:1 and a modulus of 0.4 MPa, wrapped around the periphery of the first component group at a 45° helix angle, responsible for generating bending and torsion during contraction, simulating the combined flexion and supination movements of the biceps brachii; the third component group (intermediate simulation) consists of 15 medium-strength fibers with a draw ratio of 3:1 and a modulus of 0.7 MPa, distributed in an embedded structure within the first component group (three embedded strips evenly arranged circumferentially), serving as a mechanical transition layer. The three components are co-cured to form an integrated composite with a total diameter of approximately 3 mm. Driven testing showed that when only the first component group was stimulated, the complex produced 10% axial contraction, simulating rapid elbow flexion. When the first and second component groups were stimulated simultaneously, axial contraction combined with bending and torsion, producing 30° flexion accompanied by 15° supination, simulating the main function of the biceps brachii. When the third component group was additionally stimulated, the stiffness of the intermediate layer increased, increasing the flexion speed of the complex by 20% under the same voltage while reducing impact, simulating the modulating effect of intermediate fibers on motion smoothness. Implanting this complex into a bionic arm model enabled continuous motion control from rapid punching to slow flexion and extension, with the motion trajectory highly similar to the physiological curve recorded by electromyography of the human biceps brachii.

[0350] For all flexible biomimetic muscle composites prepared by integrated molding processes (including but not limited to co-extrusion, casting, winding, hot pressing, etc.) in this application, a general electrode setting scheme can be adopted to set the electrodes in order to achieve independent electrode control of the first component group and the second component group.

[0351] In some embodiments, the motor configuration scheme can achieve electrode insulation and independent addressing through different technical paths based on the integrated molding structural characteristics.

[0352] In one embodiment, after the flexible biomimetic muscle composite is formed, electrodes are respectively set in specific areas on the surface of the component using mask protection, local coating, or micromachining techniques. The electrode setup can be achieved through mask protection or laser etching, as detailed below:

[0353] Masking method: After the composite is formed, the surface is first cleaned and dried. Then, a first mask is applied and patterned to expose the coating area of ​​the first component group. A conductive material is coated on this area and cured to form the first electrode. The first mask is then peeled off. Next, a second mask is applied and patterned to expose the coating area of ​​the second component group, while simultaneously covering and protecting the formed first electrode. A conductive material is coated on the exposed area and cured to form the second electrode. Finally, the second mask is peeled off, resulting in two sets of independent, insulated electrodes on the surface of the composite. Wires can be connected to the electrode ends as needed. Driving circuits are connected to the first and second component groups through these independent electrode lines. By applying a preset driving voltage, an electric field is applied to the corresponding component groups, exciting them to produce corresponding active deformations.

[0354] Laser etching method: First, a uniform conductive layer is coated onto the entire surface of the composite (e.g., by dip coating or spray coating). Then, using precision laser etching equipment, insulating trenches are etched along the projection interface of the two components on the surface, dividing the continuous conductive layer into two independent electrode regions that are insulated from each other. The etching depth is controlled to remove the conductive layer just enough to remove the component substrate without damaging it, and the etching width is 50-200 μm to ensure reliable insulation.

[0355] For example, taking a co-extruded core-skin composite (the first component group is the core layer, and the second component group is the skin layer), the surface of the composite after molding is a continuous skin layer material. A mask protection method is used: first, photoresist is coated on the component surface; windows are formed at corresponding positions at both ends of the core layer through exposure and development; silver paste is sprayed to form the electrodes of the first component group; after removing the photoresist, it is recoated, and windows are formed on the skin layer surface (avoiding the already coated areas); carbon grease is sprayed to form the electrodes of the second component group. The distance between the two sets of electrodes is approximately 0.5 mm, and surface resistance testing shows an insulation resistance greater than 10 MΩ, meeting the requirements for independent drive. Taking a wound composite (the first component group is the core layer, and the second component group is a spiral wound layer), the surface after molding is a wound layer material. A laser etching method is used: a carbon nanotube conductive layer is integrally coated on the surface; then, an infrared laser is used to etch a 100 μm wide annular trench along the boundary line between the wound layer and the core layer, dividing the conductive layer into a core layer electrode area (end) and a wound layer electrode area (outer periphery), with good insulation between the two.

[0356] In another embodiment, a process can be implemented that allows for the introduction of additional conductive elements during the molding process. Specifically, electrode material can be pre-embedded as an independent structure within the component or between layers during the ...

Claims

1. A flexible biomimetic muscle composite, characterized in that, include: Multiple strip-shaped components, the strip-shaped components being made of an electrically actuated material capable of active deformation under electric field excitation, and having anisotropic deformation behavior under electric field excitation at least; The multiple strip-shaped components are divided into multiple component groups, each component group including at least one strip-shaped component, and the multiple component groups include at least a first component group and a second component group, wherein the axial stiffness of the first component group is higher than that of the second component group. The first component group and the second component group are coupled through a preset force transmission structure, so that when the electric field is excited, the first driving force generated by the first component group including the first deformation and the second driving force generated by the second component group including the second deformation cooperate with each other in space, so as to realize the motion output of the flexible biomimetic muscle complex that simulates the movement characteristics of biological muscles. The first component group and the second component group are each controlled by independent electrode circuits. In response to electric field excitation with different start times and / or different durations, the contraction deformation of the first component group and the second component group forms a preset sequential or superimposed relationship in time, thereby achieving smooth establishment and attenuation of contraction force during the axial contraction of the composite; or the first component group and the second component group are arranged in an antagonistic pair in space, so that when the two are excited at the same time, the flexible bionic muscle composite produces a rigidity enhancement effect.

2. The flexible biomimetic muscle composite according to claim 1, characterized in that, The difference in axial stiffness between the first component group and the second component group is achieved by the ratio of the elastic modulus of the first component group to the elastic modulus of the second component group being greater than 3:1 and less than 20:

1.

3. The flexible biomimetic muscle composite according to claim 1, characterized in that, The difference in axial stiffness between the first component group and the second component group is achieved through the different cross-sectional geometric dimensions of the strip-shaped components in the first component group and the second component group.

4. The flexible biomimetic muscle composite according to claim 3, characterized in that, The differences in cross-sectional geometry between the strip-shaped components of the first component group and the second component group include: The ratio of the cross-sectional diameter of the strip member in the first component group to that in the second component group is greater than 1.2:1 and less than 3:

1.

5. The flexible biomimetic muscle composite according to claim 1, characterized in that, The difference in axial stiffness between the first component group and the second component group is achieved by the different products of the elastic modulus and cross-sectional area of ​​the strip-shaped components in the first component group and the second component group.

6. The flexible biomimetic muscle composite according to claim 1, characterized in that, The difference in axial stiffness between the first component group and the second component group is achieved by the difference in length of the strip-shaped components in the first component group and the second component group.

7. The flexible biomimetic muscle composite according to claim 1, characterized in that, The difference in axial stiffness between the first component group and the second component group causes the first component group to undergo a first deformation under electric field excitation, the first deformation including at least contraction motion, and causes the second component group to undergo a first contraction deformation or a second contraction deformation along its axial direction under electric field excitation. Wherein, the first contraction deformation is an auxiliary contraction deformation relative to the first deformation of the first component group, and the second component group serves as a supplementary power source for the movement of the first component group. The second contraction deformation is a counter-contraction deformation opposite to the first deformation direction, so that the second component group and the first component group form an internal force antagonism.

8. The flexible biomimetic muscle composite according to claim 1, characterized in that, The first deformation includes at least a contraction motion.

9. The flexible biomimetic muscle composite according to claim 8, characterized in that, The first deformation and the second deformation work together to produce torsional motion, spiral motion, oscillating motion or wave motion.

10. The flexible biomimetic muscle composite according to claim 8, characterized in that, The first component group and the second component group respond to electric field excitation at different start times or for different durations, so that the first deformation and the second deformation form a preset phase relationship or sequence relationship in time, thereby enabling the flexible bionic muscle complex to generate the motion output.

11. The flexible biomimetic muscle composite according to claim 1, characterized in that, The strip-shaped components constituting the first component group and / or the second component group are configured to be independently addressable and driven, and the motion output is achieved by applying differentiated driving electric fields to different strip-shaped components or different component groups.

12. The flexible biomimetic muscle composite according to claim 1, characterized in that, The first component group and the second component group have a preset first spatial structural relationship, which includes at least one of the following: First attachment type: At least one strip-shaped member in the second component group is attached longitudinally to the surface of at least one strip-shaped member in the first component group; First covering type: At least one strip-shaped component in the second component group is wrapped around the outer periphery of at least one strip-shaped component in the first component group in a spiral or woven manner; First Embedding: At least one strip-shaped member in the second component group is embedded inside at least one strip-shaped member in the first component group to form a core-skin structure; First intermittent arrangement: The first component group and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction.

13. The flexible biomimetic muscle composite according to claim 7 or 8, characterized in that, The strip-shaped components in the first component group and the strip-shaped components in the second component group are arranged in three-dimensional space along a preset spiral path. The strip-shaped components in the first component group are arranged along the main direction of the spiral to provide axial contraction constraint, and the strip-shaped components in the second component group are arranged in a cross spiral to guide torsional motion, so that the composite generates a torsional torque about its axis under electric field excitation.

14. The flexible biomimetic muscle composite according to claim 1, characterized in that, At least a portion of the strip-shaped members in the second component group are made of conductive polymer, and their driving voltage is lower than that of the first component group.

15. The flexible biomimetic muscle composite according to claim 1, characterized in that, The plurality of component groups further includes a third component group, the elastic modulus of which is greater than that of the second component group and less than that of the first component group; the first component group, the third component group, and the second component group have a preset second spatial structural relationship, the second spatial structural relationship including at least one of the following: Second attachment type: The third component group is attached to the first component group, and the second component group is attached to the third component group; Second encapsulation type: The third component group encapsulates the first component group, and the second component group encapsulates the third component group; Second Embedding: The third component group is embedded inside the first component group, and the second component group is embedded inside the third component group; Second interval arrangement: The first component group, the third component group, and the second component group are arranged alternately in a direction perpendicular to the longitudinal direction.

16. The flexible biomimetic muscle composite according to claim 1, characterized in that, The elastic moduli of the first component group, the third component group, and the second component group are distributed in a continuous gradient along the radial direction on the cross-section of the composite.

17. The flexible biomimetic muscle composite according to claim 1, characterized in that, It also includes a flexible strain sensor integrated thereon, used to monitor the deformation of the composite in real time and transmit feedback signals to the control system.

18. The flexible biomimetic muscle composite according to claim 1, characterized in that, At least a portion of the multiple strip-shaped components are integrated drive and sensing components, and the integrated drive and sensing component specifically includes at least one of the following: The first driving sensing component is composed of piezoelectric material and electro-actuating material. It is configured to actively deform under electric field excitation and collect piezoelectric signals generated by deformation in real time. The second driving sensing component is made of ion-type electro-actuated material and has an electrode array inside or on its surface. It is configured to monitor the change in ion concentration distribution in real time when active deformation occurs under electric field excitation, which is used to represent the deformation state of the component.

19. The flexible biomimetic muscle composite according to claim 1, characterized in that, The strip-shaped component itself has a multi-level structure, including: The first-level structure includes nanoscale myofibril biomimetic units composed of directionally aligned functional fillers or molecular chains; and The second-level structure includes a micron-sized myofibril biomimetic unit formed by a bundle of multiple myofibril biomimetic units, wherein the micron-sized myofibril biomimetic unit constitutes the basic unit of the first component group or the second component group.

20. The flexible biomimetic muscle composite according to claim 1, characterized in that, The strip-shaped member includes reinforcing structures disposed at both ends, the reinforcing structures including at least one of the following: The locally thickened structure has a cross-sectional area at both ends of the strip-shaped member that is larger than that in the middle region. The gradient modulus structure has a higher elastic modulus at both ends than in the middle, forming a mechanical gradient that increases from the middle to both ends. A fiber-reinforced layer, consisting of a high-strength fiber-reinforced layer or a mesh-like constraint layer covering the outer periphery of the two ends of the strip-shaped member; An anchoring interface reinforcement layer is provided at both ends of the strip-shaped member, and the interface reinforcement layer is integrally formed with the force transmission structure.

21. The flexible biomimetic muscle composite according to claim 1, characterized in that, The preset force transmission structure includes a biomimetic arrangement configuration, which is a biomimetic feather configuration. The first component group is arranged along the functional axis of the composite, and the second component group is arranged at a preset angle to the functional axis. The range of the preset angle is consistent with the physiological range of the feather angle of the muscle fibers of the target pennate or semi-pennate muscle.

22. The flexible biomimetic muscle composite according to claim 1, characterized in that, The first component group and / or the second component group are divided into multiple independently replaceable fiber modules. Each fiber module is composed of a bundle of strip-shaped components with a preset elastic modulus value, and the ends of the fiber modules are provided with mechanical interfaces and electrical connection interfaces. The fiber module is assembled in a detachable manner through the mechanical interface. The electrical connection interface is used to provide a driving electric field for the strip-shaped component within the fiber module and to transmit sensing signals when the strip-shaped component has a sensing function. By selecting fiber modules with different modulus values ​​for combination, or changing the radial arrangement order of the fiber modules on the cross-section of the flexible bionic muscle composite, the modulus spatial distribution of the flexible bionic muscle composite can be adjusted, thereby reconstructing its macroscopic motion output characteristics.

23. A flexible drive control method, characterized in that, The method, applied to the flexible biomimetic muscle complex according to any one of claims 1-22, comprises: The first component group responds to the first electric field excitation and triggers the generation of a first driving force including a first deformation; The second component group responds to the excitation of the second electric field and triggers the generation of a second driving force including a second deformation; Based on the force transmission structure coupling between the first component group and the second component group, the first driving force and the second driving force cooperate with each other in space, so that the flexible biomimetic muscle complex simulates the motion output that conforms to the movement characteristics of biological muscles.

24. A driving module, characterized in that, include: Flexible substrate; as well as At least one flexible biomimetic muscle complex as described in any one of claims 1-22, wherein the flexible biomimetic muscle complex is fixed on the flexible substrate.

25. A biomimetic robot, characterized in that, The bionic robot integrates at least two drive modules as described in claim 24.

Images