Exoskeleton driving system with multi-mode torque coupling and intelligent brain cooperative control

Abstract

The invention relates to the technical field of exoskeleton robots, in particular to an exoskeleton driving system with multi-mode torque coupling and intelligent-brain cooperative control, which comprises a flexible torque coupling mechanism, a holographic intelligent-brain cooperative control strategy, an energy safety integrated shield module and an ultra-light integrated magic body structure, wherein the flexible torque coupling mechanism is used for adapting to power and flexibility of different scenes and carrying out torque output and rigidity continuous adjustment; a holographic intelligent brain cooperative control strategy is fused with perception information, data are analyzed through quantum federal learning and a bionic algorithm, torque is optimized, and cooperative control is carried out; the energy and safety integrated shield module is used for constructing a double-safety system, and recycling is achieved through magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage. The ultra-light integrated magic body structure is matched with the human body size to adjust the exoskeleton, and the man-machine cooperation response is improved through the composite carbon fiber material and the highly integrated design. Therefore, the problems of high energy consumption, complex mechanical structure, insufficient stability and the like in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the field of exoskeleton robot technology, and specifically to an exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control. Background Art

[0002] As an emerging equipment integrating mechanical engineering, biomedicine and intelligent control, exoskeleton robots have demonstrated important application value in rehabilitation medicine, industrial load-bearing and military individual enhancement. Their core drive system needs to achieve efficient coupling of human and machine motion, that is, accurately matching the torque and motion requirements of human joints in various motion modes such as walking, running, going up and down stairs, and climbing slopes, while taking into account wearing comfort, motion safety and energy utilization efficiency. With the expansion of application scenarios, exoskeleton drive systems face technical challenges such as dynamic torque distribution under multi-modal motion, real-time perception of human motion intentions, and environmental adaptability control. It is urgent to achieve innovative breakthroughs in mechanical transmission mechanisms and intelligent control algorithms to improve human-machine collaborative performance and reliability under complex working conditions.

[0003] However, traditional exoskeleton drive technology is limited by the inherent defects of mechanical transmission and control algorithms: at the mechanical level, the fixed-speed ratio reducer and single motor design lead to inefficiency during multi-modal switching (such as motor overload when climbing and response lag when running), and the system's own weight and load capacity are unbalanced; at the control level, the accuracy of motion intention recognition relying on single signals such as encoders and plantar pressure sensors is insufficient, and traditional PID and other control algorithms result in torque control errors of more than ±15% due to parameter mismatch of the human body dynamics model, and lack environmental perception capabilities; at the energy management level, the energy recovery rate under negative working conditions is less than 20%, and the battery life is generally less than 2 hours, which makes it difficult to meet the long-term operation requirements of industrial and military scenarios. The overall technology faces multiple challenges of multi-modal adaptability, human-machine collaboration accuracy and energy efficiency. Summary of the Invention

[0004] This application provides an exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control to solve the problems of high energy consumption, complex mechanical structure and insufficient stability in the existing technology.

[0005] The first embodiment of the present application provides an exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control, including: a smart torque coupling mechanism, a holographic intelligent brain collaborative control, an energy-safety integrated shield module, and an ultra-lightweight integrated phantom structure; wherein, the smart torque coupling mechanism is used to adapt to the power and flexibility in different complex motion scenes, and through the coordination of the planetary-magnetoelectric-fluidic integrated transmission unit and the internal implicit harmonic deceleration integrated structure, 0-600N·m torque output and 5-50N·m / rad stiffness continuous adjustment are performed; the holographic intelligent brain collaborative control is used to integrate It combines sensory information to provide intelligent motion assistance in complex motion scenarios, combines quantum federated learning and bionic neural control algorithms, deeply analyzes motion intentions and environmental data, optimizes torque strategies and performs dynamic collaborative control; the energy-safety integrated shield module is used to build a dual safety assurance system, and recycles energy through magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage; the ultra-lightweight integrated phantom structure is used to adapt to different human body sizes and enhance human-machine collaborative motion. Through multi-dimensional composite carbon fiber materials and highly integrated design, lightweight adjustment and high-speed response of exoskeleton joints are achieved.

[0006] Preferably, the dynamic torque coupling mechanism includes a planetary-magnetoelectric-fluidic integrated transmission unit and an implicit harmonic deceleration integrated structure, wherein the planetary-magnetoelectric-fluidic integrated transmission unit is used to integrate planetary gear rigid transmission, magnetic powder clutch dynamic torque coupling and magnetorheological fluid flexible stiffness adjustment technology, and through the coordinated control of the three-stage planetary gear set and the magnetorheological fluid damping unit, the wide-range torque output and stiffness continuous adaptive adjustment are performed; the implicit harmonic deceleration integrated structure is used to embed the ultra-precision harmonic reducer in the motor stator, combined with the shape memory alloy wire wound flexible wheel collaborative design, to perform dynamic coupling control of stiffness.

[0007] Preferably, the holographic intelligent brain collaborative control includes a dual-loop intelligent adjustment system and a bionic neural-myoelectric fusion control algorithm, wherein the dual-loop intelligent adjustment system is used to integrate the current-speed dual closed-loop control architecture and the quantum federated learning dynamic decision-making mechanism, and performs gait recognition, stiffness adjustment and power distribution through inner-loop real-time torque compensation and outer-loop energy-safety collaborative optimization; the bionic neural-myoelectric fusion control algorithm is used to integrate biological electromyographic signal real-time decoding and bionic central pattern generator dynamic modeling technology, and construct a three-level mapping model of electromyographic signal-neural intention-motor instruction through deep coupling of the surface electromyography sensor array and the LSTM neural network to perform real-time compensation of exoskeleton joint torque.

[0008] Preferably, the formula of the quantum federated learning dynamic decision-making mechanism is: ; in, is the updated global model parameter; is the global model parameter before updating; is the learning rate; N is the total number of nodes participating in federated learning; i, j are index variables; λ is a hyperparameter; is the Jensen-Shannon divergence; is the data distribution of the jth node; is the global data distribution; is an exponential function; is an operator for finding the gradient of model parameters; is the quantum state associated with the model parameters; is the quantum Hamiltonian; is the expectation value of the quantum Hamiltonian in the quantum state.

[0009] Preferably, the energy-safety integrated shield module includes a magnetic flow-elastic dual-effect energy management module and a Starlink encryption-neural reasoning security module, wherein the magnetic flow-elastic dual-effect energy management module is used to convert motion kinetic energy into electrical energy in real time, and release elastic potential energy to assist joint movement during the gait swing phase; the Starlink encryption-neural reasoning security module is used to perform nanosecond-level secure evidence storage and real-time verification of control instructions, combined with the 1D-CNN+LSTM neural network self-evolution learning of complex fault modes, and equipped with an independent hardware security chip for millisecond-level braking response, to build a triple safety protection system of encryption evidence-intelligent diagnosis-hardware redundancy.

[0010] Preferably, the ultra-lightweight integrated phantom structure includes a multi-dimensional composite carbon fiber lightweight module and a high-speed integrated execution module, wherein the multi-dimensional composite carbon fiber lightweight module is used to dynamically adjust the axial stiffness and vibration attenuation during high-speed movement, and adopts a three-dimensional woven carbon fiber composite material, combined with an adjustable joint mechanism formed by 3D printing, to improve movement flexibility and adapt to the human body shape under the ultra-lightweight design of a single joint mass of less than 1.3 kg; the high-speed integrated execution module is used for nanosecond synchronous acquisition of sensor data, millisecond-level solution of torque instructions and sub-millisecond precision drive of magnetorheological fluid / magnetic powder clutch, integrating a 2.5Mbps CANFD bus high-speed data link, a 22ms supercomputing edge node and a 25kHz ultra-high frequency PWM drive technology to provide instant power response for high-speed movement of humans and machines.

[0011] The second embodiment of the present application provides an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control, including: acquiring motion scene data; according to the motion scene data, adjusting the joint stiffness through the collaborative control of the planetary gear set, the magnetoelectric drive unit and the rheological damper, and generating basic dynamic parameters suitable for scenes such as walking, running, jumping, and weight bearing; inputting the basic dynamic parameters into the control algorithm of integrated quantum federated learning, combining the biodynamic model, the real-time mapping of the human joint angle and muscle electrical signals, parsing the motion intention through the intelligent brain and generating a torque compensation strategy, and reversely driving the rheological damper to adjust the stiffness. The speed change rate and the output duty cycle of the magnetoelectric unit are compared; according to the torque compensation strategy, the kinetic energy is recovered through the magnetorheological fluid brake device and the shape memory alloy spring, and stored in a 100mF supercapacitor group after DC-DC conversion; based on the stable power supply of the supercapacitor group, the adaptive wheelbase adjustment mechanism of the multi-dimensional woven carbon fiber frame is used to match the wearer's leg length, and combined with the integrated joint servo mechanism, the power parameters are optimized, and the optimized power parameters are converted into joint motion torque. At the same time, the pressure distribution of the contact surface is dynamically adjusted through the pressure sensor array to perform human-machine collaborative motion and reduce the error rate.

[0012] The third aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and runnable on the processor. The processor executes the program to implement an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control as in the above embodiment.

[0013] The fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control as described in the above embodiment.

[0014] The fifth embodiment of the present application provides a computer program product, including a computer program or instructions, for implementing an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control as in the above embodiment.

[0015] Therefore, this application has the following beneficial effects: The embodiment of the present application uses a flexible torque coupling mechanism with the help of a planetary-magnetoelectric-fluidic and harmonic deceleration integrated structure, which can be continuously adjusted within the torque range of 0-600N·m and the stiffness range of 5-50N·m / rad, taking into account the power output of heavy-duty operations and the flexibility requirements of fine movements, and adapting to complex motion scenes; the holographic intelligent brain collaborative control strategy uses quantum federated learning and bionic neural control algorithms to deeply integrate multi-source perception information, accurately analyze the user's motion intentions and environmental data, and then dynamically optimize the torque strategy, so that the exoskeleton and human body movements are highly coordinated and can be used in complex scenes. The integrated energy-safety shield module utilizes magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to create an energy recycling system and dual safety guarantees. This system not only recovers kinetic energy to improve battery life, but also reduces impact risk through potential energy buffering. The ultra-lightweight integrated phantom structure utilizes multi-dimensional composite carbon fiber and a highly integrated design to achieve lightweight exoskeleton joints (weight reduction of over 30%) and high-speed response (joint movement delay <5ms), adapting to users of different body types and enhancing the comfort and flexibility of human-machine collaborative motion. This addresses the high energy consumption, complex mechanical structure, and insufficient stability inherent in existing technologies.

[0016] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which: Figure 1 This is a schematic diagram of the structure of an exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control according to an embodiment of the present application; Figure 2 A schematic diagram of a flexible torque coupling mechanism according to one embodiment of the present application; Figure 3 A schematic diagram of an exoskeleton transmission system according to one embodiment of the present application; Figure 4 A schematic diagram of holographic intelligent brain collaborative control according to one embodiment of the present application; Figure 5 A schematic diagram of an energy-safety integrated shield module provided according to one embodiment of the present application; Figure 6 A schematic diagram of an ultra-lightweight integrated phantom structure provided according to one embodiment of the present application; Figure 7 A schematic diagram of a civilian rehabilitation exoskeleton provided according to one embodiment of the present application; Figure 8This is a flow chart of a logistics warehousing exoskeleton system provided according to one embodiment of the present application; Figure 9 This is a flow chart of an exoskeleton driving method with multimodal torque coupling and intelligent brain collaborative control according to an embodiment of the present application; Figure 10 A schematic diagram of an emergency rescue exoskeleton system according to one embodiment of the present application; Figure 11 This is a flow chart of an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control according to one embodiment of the present application;

[0018] Figure 12 A schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0020] The following describes an exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control according to an embodiment of the present application with reference to the accompanying drawings. In response to the problem of insufficient stability mentioned in the above background technology, the present application provides an exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control. In this system, through the flexible torque coupling mechanism with the help of planetary-magnetoelectric-fluid body and harmonic deceleration integrated structure, it can be continuously adjusted within the torque range of 0-600N·m and the stiffness range of 5-50N·m / rad, taking into account the power output of heavy-load operations and the flexibility requirements of fine movements, and adapting to complex motion scenes; holographic intelligent brain collaborative control uses quantum federated learning and bionic neural control algorithms to deeply integrate multi-source perception information, accurately analyze the user's movement intentions and environmental data, and then Dynamically optimized torque strategies enable highly coordinated exoskeleton and human motion, providing intelligent and efficient motion assistance even in complex scenarios. The integrated energy-safety shield module utilizes magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to create an energy recycling system and dual safety guarantees. This not only recycles motion energy to improve battery life, but also reduces impact risk through potential energy buffering. The ultra-lightweight integrated phantom structure utilizes multi-dimensional composite carbon fiber and a highly integrated design, achieving lightweight exoskeleton joints (over 30% weight reduction) and high-speed response (joint motion delay <5ms), adapting to users of varying body shapes and enhancing the comfort and flexibility of human-machine collaborative motion. This addresses the challenges of high energy consumption, complex mechanical structures, and insufficient stability inherent in existing technologies.

[0021] Figure 1 A schematic diagram of the composition of an exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control provided in an embodiment of the present application.

[0022] The present application provides an exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control. The system 10 includes: Flexible torque coupling mechanism 100, holographic intelligent brain collaborative control 200, energy-safety integrated shield module 300, ultra-lightweight integrated phantom structure 400.

[0023] Among them, the flexible torque coupling mechanism is used to adapt to the power and flexibility in different complex motion scenarios. Through the coordination of the planetary-magnetoelectric-rheological fluid integrated transmission unit and the implicit harmonic reduction integrated structure, 0-600N・m torque output and 5-50N・m / rad stiffness are continuously adjusted; holographic intelligent brain collaborative control is used to integrate perception information and provide intelligent motion assistance in complex motion scenarios. Combined with quantum federated learning and bionic neural control algorithms, it deeply analyzes motion intentions and environmental data, optimizes torque strategies and performs dynamic collaborative control; the energy-safety integrated shield module is used to build a dual safety protection system, and energy is recycled through magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage; the ultra-lightweight integrated phantom structure is used to adapt to different human body sizes and enhance human-machine collaborative motion. Through multi-dimensional composite carbon fiber materials and highly integrated design, lightweight adjustment and high-speed response of exoskeleton joints are achieved.

[0024] It can be understood that in the embodiment of the present application, the flexible torque coupling mechanism can be used with the help of the planetary-magnetoelectric-fluid body and the harmonic deceleration integrated structure to continuously adjust the torque range of 0-600N·m and the stiffness range of 5-50N·m / rad, taking into account the power output of heavy-duty operations and the flexibility requirements of fine movements, and adapting to complex motion scenes; the holographic intelligent brain collaborative control strategy uses quantum federated learning and bionic neural control algorithms to deeply integrate multi-source perception information, accurately analyze the user's movement intentions and environmental data, and then dynamically optimize the torque strategy, so that the exoskeleton and human body movements are highly coordinated. It provides intelligent and efficient motion assistance even in complex scenarios. The integrated energy-safety shield module utilizes magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to create an energy recycling system and dual safety guarantees. This not only recycles motion energy to improve battery life, but also reduces impact risk through potential energy buffering. The ultra-lightweight integrated phantom structure utilizes multi-dimensional composite carbon fiber and a highly integrated design to achieve lightweight exoskeleton joints (weight reduction of over 30%) and high-speed response (joint movement delay <5ms), adapting to users of different body shapes and enhancing the comfort and flexibility of human-machine collaborative motion. This solves the problems of high energy consumption, complex mechanical structure, and insufficient stability in existing technologies.

[0025] In the embodiment of the present application, the dynamic torque coupling mechanism 100 further includes: Figure 2 As shown, the planetary-magnetoelectric-fluidic integrated transmission unit and the internal implicit harmonic reduction integrated structure.

[0026] Among them, the planetary-magnetoelectric-fluidic integrated transmission unit is used to integrate planetary gear rigid transmission, magnetic powder clutch dynamic torque coupling and magnetorheological fluid flexible stiffness adjustment technology. Through the coordinated control of the three-stage planetary gear set and the magnetorheological fluid damping unit, wide-range torque output and continuous adaptive adjustment of stiffness are performed; the implicit harmonic reduction integrated structure is used to embed the ultra-precision harmonic reducer into the motor stator, combined with the collaborative design of the shape memory alloy wire-wound flexible wheel to perform dynamic coupling control of stiffness.

[0027] It can be understood that the embodiment of the present application integrates rigid transmission, dynamic torque coupling and flexible stiffness adjustment technologies through the planetary-magnetoelectric-rheological fluid integrated transmission unit, and cooperates with the three-stage planetary gear set and the magnetorheological fluid damping unit to perform wide-range torque output and continuous adaptive adjustment of stiffness to meet the power and flexibility requirements of complex scenarios; the implicit harmonic reduction integrated structure embeds the ultra-precision harmonic reducer into the motor stator, combined with the shape memory alloy wire-wound flexible wheel design, to perform dynamic coupling control of stiffness, thereby improving the integration and control accuracy of the transmission system.

[0028] For example, Figure 3 As shown, in medical rehabilitation exoskeleton robots, planetary-magnetoelectric-fluidic integrated transmission units can achieve coordinated adjustment of power and flexibility in complex motion scenarios. For example, the intelligent exoskeleton joint module developed by the University of Science and Technology of China uses the coordinated control of a three-stage planetary gearset and a magnetorheological fluid damping unit. When carrying heavy objects, the planetary gears provide a wide torque output range of 0-600 N·m, while the magnetorheological fluid adjusts the stiffness by 5-50 N·m / rad in real time based on the load, ensuring joint stability under high loads. In rehabilitation training, the magnetic powder clutch uses dynamic torque coupling technology combined with the flexible adjustment properties of magnetorheological fluid to enable the exoskeleton to not only precisely assist joint movement, but also achieve dynamic coupled stiffness control through the coordinated design of a shape memory alloy wire-wound flexspline, improving the flexibility of human-machine interaction.

[0029] In the embodiment of the present application, the holographic intelligent brain collaborative control 200 includes: Figure 4 As shown, dual-loop intelligent adjustment system and bionic neural-myoelectric fusion control algorithm.

[0030] Among them, the dual-loop intelligent adjustment system is used to integrate the current-speed dual closed-loop control architecture and the quantum federated learning dynamic decision-making mechanism, and performs gait recognition, stiffness adjustment and power distribution through the inner-loop real-time torque compensation and outer-loop energy-safety collaborative optimization; the bionic neural-myoelectric fusion control algorithm is used to integrate the real-time decoding of biological electromyographic signals and the dynamic modeling technology of the bionic central pattern generator, and through the deep coupling of the surface electromyographic sensor array and the LSTM neural network, construct a three-level mapping model of electromyographic signals-neural intentions-motor commands to perform real-time compensation of exoskeleton joint torque.

[0031] It can be understood that the embodiment of the present application integrates dual closed-loop control and quantum learning through a dual-loop intelligent adjustment system, uses inner-loop real-time torque compensation to ensure accurate power output during the movement process, and uses outer-loop energy-safety collaborative optimization to optimize the global strategy of gait recognition, stiffness adjustment and power distribution, thereby improving control stability in complex scenarios; the bionic neural-myoelectric fusion control algorithm uses the deep coupling of the surface electromyography sensor array and the LSTM neural network to construct a three-level mapping model from electromyographic signal acquisition, neural intention analysis to motion command generation, thereby realizing real-time dynamic compensation of exoskeleton joint torque and accurately capturing human body movement intentions.

[0032] For example, in a lower-limb rehabilitation exoskeleton, the dual-loop intelligent adjustment system demonstrates precise adaptation to complex gaits. When the wearer performs various movements, such as walking on level ground and ascending and descending stairs, the system monitors joint torque fluctuations in real time through an inner-loop current-velocity dual closed-loop control system. To address the transient high load demands during stair climbing, the inner loop immediately activates a torque compensation mechanism to ensure a stable output of 300 N·m of peak torque at the knee joint. Simultaneously, the outer loop dynamically analyzes the gait cycle using quantum federated learning. During level ground walking, the stiffness is adjusted to 15 N·m / rad to match the human body's natural gait, and automatically increases to 35 N·m / rad during squatting to ensure joint support. This system also optimizes energy distribution and extends battery life by 20%. Tests on a domestic exoskeleton product have shown that, when equipped with this system, the joint angle error during walking for spinal cord injury patients has been reduced from ±8° to ±3°. Furthermore, in the event of a sudden stumble, the outer loop's energy-safety synergy module triggers a sudden increase in stiffness within 50ms, reducing the risk of falls by 60%.

[0033] It should be noted that the formula for the dynamic decision-making mechanism of quantum federated learning is: ; in, is the updated global model parameter; is the global model parameter before updating; is the learning rate; N is the total number of nodes participating in federated learning; i, j are index variables; λ is a hyperparameter; is the Jensen-Shannon divergence; is the data distribution of the jth node; is the global data distribution; is an exponential function; is an operator for finding the gradient of model parameters; is the quantum state associated with the model parameters; is the quantum Hamiltonian; is the expectation value of the quantum Hamiltonian in the quantum state.

[0034] In the embodiment of the present application, the energy-safety integrated shield module 300 includes: Figure 5 As shown, the magnetic flow-elastic dual-effect energy management module and the Starlink encryption-neural reasoning security module.

[0035] Among them, the magnetofluid-elastic dual-effect energy management module is used to convert kinetic energy into electrical energy in real time, and release elastic potential energy to assist joint movement during the gait swing phase; the Starlink encryption-neural reasoning security module is used to perform nanosecond-level secure storage and real-time verification of control instructions, combined with the 1D-CNN+LSTM neural network self-evolution to learn complex fault modes, and equipped with an independent hardware security chip for millisecond-level braking response, to build a triple security protection system of encryption storage-intelligent diagnosis-hardware redundancy.

[0036] It can be understood that the embodiment of the present application converts kinetic energy into electrical energy in real time through the magnetic flow-elastic dual-effect energy management module, releases elastic potential energy to assist joint movement during gait swing, improves energy utilization, and extends device life; the Starlink encryption-neural reasoning security module performs nanosecond-level secure storage and verification of control instructions, uses neural networks to learn fault modes, and cooperates with hardware chips to achieve millisecond-level braking response.

[0037] For example, in high-risk industrial load-bearing exoskeletons, the Starlink encryption-neural inference security module demonstrates millisecond-level response capabilities to sudden risks. When a wearer carries a 200kg load while working at height, the module first performs nanosecond blockchain authentication and two-way encryption verification on the joint motor's control commands, ensuring that the commands have not been tampered with or hijacked during transmission. If the LSTM neural network detects abnormal current fluctuations in the knee motor (such as a sudden torque surge caused by a stall) through 1D-CNN feature extraction, the system immediately triggers the intelligent diagnostic module, identifying the gearbox wear failure mode within 0.8ms and simultaneously activating an independent hardware security chip to cut off power output and lock the joint within 5ms, a 90% increase in response speed compared to traditional safety systems. Actual data from a logistics robot shows that after being equipped with this module, the safety accident rate due to control command tampering or mechanical failure dropped from 0.3 times per thousand hours to 0.02 times per thousand hours. Furthermore, even in extreme electromagnetic interference environments, the command verification accuracy remained at 99.99%.

[0038] In the embodiment of the present application, the ultra-lightweight integrated phantom structure 400 includes, for example Figure 6 As shown, it includes a multi-dimensional composite carbon fiber lightweight module and a high-speed integrated execution module.

[0039] Among them, the multi-dimensional composite carbon fiber lightweight module is used to dynamically adjust the axial stiffness and vibration attenuation during high-speed movement. It adopts three-dimensional woven carbon fiber composite materials and combines with an adjustable joint mechanism formed by 3D printing. With an ultra-lightweight design of a single joint mass of less than 1.3kg, it improves movement flexibility and adapts to the human body shape; the high-speed integrated execution module is used for nanosecond synchronous acquisition of sensor data, millisecond-level solution of torque instructions and sub-millisecond precision drive of magnetorheological fluid / magnetic powder clutch, integrating 2.5Mbps CANFD bus high-speed data link, 22ms supercomputing edge node and 25kHz ultra-high frequency PWM drive technology to provide instant power response for high-speed movement of humans and machines.

[0040] It can be understood that the embodiment of the present application uses three-dimensional woven carbon fiber composite materials and 3D printed adjustable joint mechanisms to achieve an ultra-lightweight design with a single joint mass of less than 1.3 kg. At the same time, it dynamically adjusts the axial stiffness, reduces vibration during high-speed movement, improves movement flexibility and adapts to different human body shapes; the high-speed integrated execution module relies on nanosecond sensor data synchronous acquisition, millisecond torque command solution and sub-millisecond drive technology, combined with the CANFD bus, supercomputing edge nodes and ultra-high frequency PWM drive, to provide instant power response for high-speed movement of humans and machines.

[0041] For example, Figure 7 As shown, in the field of civilian rehabilitation exoskeletons, lightweight multi-dimensional composite carbon fiber modules significantly improve wearer comfort and mobility. For example, an exoskeleton designed for gait assistance for the elderly utilizes a 3D-woven carbon fiber composite material and a 3D-printed adjustable knee joint mechanism. Each joint weighs only 1.2 kg, a 40% reduction compared to traditional aluminum alloy structures. When the wearer ascends or descends stairs, the module dynamically adjusts axial stiffness, increasing it to 25 N·m / rad during the support phase to stabilize joint loads and reducing it to 10 N·m / rad during the swing phase to reduce motion resistance. A vibration damping design reduces vibration amplitude during high-speed knee swing by 60%, effectively preventing joint strain. For users with heights ranging from 1.5 to 1.9 meters, the adjustable joint mechanism allows tool-free leg length adjustment within 5 minutes, expanding the hip flexion range to 120°-160°, significantly improving user-machine fit. Actual measured data shows that the battery life of an exoskeleton equipped with this module is extended by 35%, and the wearer's average daily usage time is increased from 4 hours to 7 hours. In addition, due to the reduced weight and adaptive stiffness adjustment, the energy consumption of walking is reduced by 28% compared to traditional steel exoskeletons.

[0042] The embodiment of the present application proposes an exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control. Through the flexible torque coupling mechanism, with the help of planetary-magnetoelectric-fluid body and harmonic deceleration integrated structure, it can be continuously adjusted in the torque range of 0-600N·m and the stiffness range of 5-50N·m / rad, taking into account the power output of heavy-load operations and the flexibility requirements of fine movements, and adapting to complex motion scenes; the holographic intelligent brain collaborative control strategy uses quantum federated learning and bionic neural control algorithms to deeply integrate multi-source perception information, accurately analyze the user's movement intention and environmental data, and then dynamically optimize the torque strategy to enable the exoskeleton to The skeleton is highly coordinated with human movement, providing intelligent and efficient motion assistance even in complex scenarios. The integrated energy-safety shield module utilizes magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to create an energy recycling system and dual safety guarantees. This not only recycles motion energy to improve battery life, but also reduces impact risk through potential energy buffering. The ultra-lightweight integrated phantom structure utilizes multi-dimensional composite carbon fiber and a highly integrated design to achieve lightweight exoskeleton joints (weight reduction of over 30%) and high-speed response (joint movement delay <5ms), adapting to users of different body shapes and enhancing the comfort and flexibility of human-machine collaborative motion. This solves the problems of high energy consumption, complex mechanical structure, and insufficient stability in existing technologies.

[0043] The following will describe an exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control through a specific embodiment. Figure 8 Shown, including: A large-scale logistics and warehousing center has introduced an exoskeleton drive system. The exoskeleton's main frame is constructed from T1000-grade high-strength carbon fiber composite material, molded in an autoclave. The joint bearings have been upgraded to ceramic ball bearings. Combined with a modular transmission unit, the exoskeleton can stably carry a load of 100 kg. The planetary gear transmission utilizes a high-precision helical planetary gearset. The center gear is directly connected to the output shaft of a high-performance servo motor (such as the Maxon EC motor), while the planetary carrier connects to the exoskeleton's joint loads. The integrated internal implicit harmonic reduction mechanism utilizes a hollow cup design. The wave generator is integrated into the hollow shaft of the sun gear, the flexspline is connected to the planetary carrier, and the rigid gear is fixed to the gearbox housing. These two components work together to achieve a torque output of 0-600 N·m. The magneto-rheological system incorporates annular magneto-rheological fluid chambers in key locations within the planetary gearbox. When carrying goods, the system automatically switches power modes according to the load: when carrying light loads (≤20kg), the intelligent brain-assisted collaborative control system reduces the current of the electromagnetic coil, reducing the viscosity of the magnetorheological fluid, adjusting the stiffness to 5N・m / rad, and activating the magnetorheological fluid damper to ensure joint flexibility; when carrying heavy loads, the current is increased to solidify the magnetorheological fluid, and the stiffness is increased to 50N・m / rad. Combined with the planetary gear and harmonic reducer, multiple components work together to ensure stable power output.

[0044] When handling cargo, the system automatically switches power modes based on load. Under light loads, magnetorheological fluid dampers are activated to ensure flexibility, while under heavy loads, multiple components collaborate to ensure stable operation. An edge computing platform is built around the NVIDIA Jetson AGX Orin, supplemented by the STM32H750 chip to create a dual-loop intelligent control system. The inner loop uses the STM32H750 chip for precise torque compensation, real-time monitoring of joint loads, and rapid adjustment of motor output. The outer loop utilizes a quantum federated learning algorithm developed within the TensorFlow framework, integrating data exchanged between the 5G network and cloud servers to dynamically optimize control parameters for energy savings and efficiency gains. The perception system has been further expanded, with high-precision miniature six-axis inertial sensors (MPU6050, ±0.1° accuracy) installed at each exoskeleton joint, and a 4×4 pressure sensor array (FlexiForce A401) positioned on the soles of the feet. High-precision cargo weight sensors have been added to the shoulders and back trays, communicating with the control module in real time via the CAN bus. Combined with a helmet-mounted lidar (Velodyne VLP-16) and a binocular camera (Logitech C922), this creates a comprehensive environmental and load perception network. At the software level, the bionic neural control algorithm utilizes a long short-term memory (LSTM) model to process time-series data collected by the sensors. Combined with environmental perception information, it can predict human motion intentions 0.5 seconds in advance and optimize torque strategies.

[0045] The kinetic energy recovery system has been upgraded to a dual-action magnetoelastic energy module. A magnetorheological fluid damper is installed at the joint's rotational axis. A piston is linked to the joint, pushing the magnetorheological fluid to cut through magnetic flux lines during movement, generating electricity. This energy is then stored in a 12Ah, 24V lithium iron phosphate battery pack via a rectifier circuit. A spring-assisted structure has been added to store and release elastic potential energy during joint flexion and extension, significantly reducing energy consumption and extending the exoskeleton's battery life to six hours. The potential energy storage unit utilizes nickel-titanium alloy wire, with multiple gold wires arranged in groups at the knee and hip joints. This energy is stored by pre-stretching, and then pulsed heating by an intelligent temperature-controlled circuit causes it to return to its original shape within one second, releasing the energy to assist joint movement. The exoskeleton is equipped with a variety of sensors to form a real-time monitoring network. Pressure sensors monitor the forces on the lumbar spine and joints, temperature sensors prevent component overheating, and an IMU monitors posture. The security assurance mechanism has been enhanced to the Starlink encryption-neural reasoning security module, which uses encryption algorithms to ensure the security of command transmission and learns fault modes through neural networks. Once an abnormality is detected, the hardware security chip triggers the electromagnetic brake within 50μs and stops the exoskeleton movement within 0.3 seconds to protect the safety of workers.

[0046] The main frame's structural design was optimized through finite element analysis, reducing the overall weight to 8kg (excluding batteries) while ensuring strength (tensile strength ≥7000MPa). The joint module utilizes a modular quick-release design, connecting to the carbon fiber frame via a dovetail slot interface and equipped with an adjustable joint mechanism. Customizable to the worker's body shape and work habits, it can be customized within 5 minutes, ensuring both comfort and flexibility. In the integrated joint design, a flat, frameless torque motor (such as the Kollmorgen AKM series) is directly embedded in the joint shaft, paired with a hollow encoder (17-bit accuracy) for real-time angular feedback. Furthermore, the sensor, driver circuit board, and transmission components are stacked in a stacked layout, reducing the joint's volume by 30% and increasing the response speed to 20ms.

[0047] In summary, the embodiments of the present application significantly improve the efficiency of logistics and warehousing operations through multi-technical integration and innovative design. The high-strength carbon fiber frame and ceramic bearings give it a stable load-bearing capacity of 100kg, and the intelligent switching power mode takes into account both light-load flexibility and heavy-load stability; the dual-loop intelligent adjustment system combines multi-source perception and intelligent algorithms to accurately predict movement intentions and achieve energy saving and efficiency improvement; the magnetic flow-elastic dual-effect energy module extends the battery life to 6 hours, and cooperates with the Starlink encryption and millisecond-level braking safety module to protect operational safety in all directions; the lightweight and modular design supports 5-minute rapid adaptation, which speeds up joint response by 30%. The system effectively reduces the labor intensity and risk of work-related injuries of workers, greatly improves the efficiency of warehousing and handling, and provides a reliable solution for the intelligent upgrade of the logistics industry.

[0048] Next, an exoskeleton driving method with multimodal torque coupling and intelligent brain collaborative control proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0049] like Figure 9 As shown, the exoskeleton driving method with multimodal torque coupling and intelligent brain collaborative control includes the following steps: In step S101 , motion scene data is acquired.

[0050] It can be understood that the embodiments of the present application accurately identify the user's exercise needs and status in different scenarios by acquiring motion scene data.

[0051] In step S102, according to the motion scene data, the joint stiffness is adjusted through the coordinated control of the planetary gear set, the magnetoelectric drive unit and the rheological damper to generate basic dynamic parameters suitable for scenes such as walking, running, jumping, and weight bearing.

[0052] Among them, the rheological damper is an intelligent damping device that uses electric current or magnetic field to act on rheological materials, causing their mechanical properties such as viscosity and shear stress to undergo reversible changes within milliseconds, thereby realizing dynamic adjustment of the damping force.

[0053] It can be understood that the embodiments of the present application control the mechanical properties of rheological materials through electric current or magnetic field, realize dynamic adjustment of damping force within milliseconds, and cooperate with planetary gear sets and magnetoelectric drive units to accurately match the flexible joint stiffness required for the natural gait of the human body in walking scenarios, thereby reducing energy consumption during exercise; in running and jumping scenarios, the damping force is quickly increased to enhance joint support and ensure movement stability. In weight-bearing scenarios, the damping characteristics are optimized in real time according to load changes, thereby improving the exoskeleton's carrying capacity and operational flexibility, effectively reducing vibration impact, and providing adaptive, efficient and stable basic power parameter adjustment for multiple sports scenarios.

[0054] For example, in medical rehabilitation scenarios, the exoskeleton used for patients with ASIA-grade CD spinal cord injuries maintains a stiffness of 15N·m / rad in the rheostatic damper when the patient walks on flat ground, perfectly matching the human body's natural gait and making walking easier for the patient. When going up and down stairs, the system identifies the step height through IMU inclination data, and the stiffness of the rheostatic damper rapidly increases to 25N·m / rad. At the same time, the dual-loop intelligent adjustment system increases the knee joint torque to 350N·m, providing stable joint support for the patient and ensuring that he or she can complete the stairs climbing and descending movements safely and smoothly.

[0055] In step S103, the basic dynamic parameters are input into the control algorithm of integrated quantum federated learning, combined with the biodynamic model, real-time mapping of human joint angles and muscle electrical signals, and the intelligent brain analyzes the movement intention and generates a torque compensation strategy, which reversely drives the rheological damper to adjust the stiffness change rate and the output duty cycle of the magnetoelectric unit.

[0056] Among them, the torque compensation strategy is a control strategy in which the exoskeleton drive system dynamically adjusts the joint torque through component collaboration based on data such as load and human movement intention, compensates for torque deviations caused by changes in load, terrain or movement state, thereby providing stable and adaptive power support in scenarios such as walking and carrying weight, and improving movement efficiency and safety.

[0057] It can be understood that the embodiments of the present application use a torque compensation strategy to enable the exoskeleton drive system to dynamically adjust the joint torque through component collaboration based on data such as load and human movement intention, effectively compensate for torque deviations caused by load fluctuations, terrain changes or changes in human movement state, and provide users with stable and adaptive power support in scenarios such as walking and carrying weight, thereby improving the movement efficiency and safety of the exoskeleton in different movement scenarios.

[0058] For example, Figure 10As shown in the figure, in emergency rescue scenarios, when rescuers wear exoskeleton equipment equipped with a torque compensation strategy and enter rugged ruins to carry heavy objects, the system can sense load changes such as ground slope and weight of heavy objects in real time, as well as rescuers' movement intentions such as lifting legs and bending over. Through the coordination of planetary gear sets, magnetoelectric drive units and rheological dampers, the system dynamically adjusts the joint output torque to compensate for the torque deviation caused by terrain and load, making rescuers walk more steadily and carry more labor-saving, which not only ensures the efficiency of rescue operations, but also reduces the risk of injury due to excessive fatigue.

[0059] In step S104, according to the torque compensation strategy, the kinetic energy is recovered through the magnetorheological fluid brake device and the shape memory alloy spring, and is stored in the 100mF supercapacitor bank after DC-DC conversion.

[0060] Among them, the shape memory alloy spring is a special spring made of shape memory alloy material, which can automatically return to a preset shape and generate elastic deformation force when the temperature or stress changes.

[0061] It can be understood that the embodiment of the present application efficiently recovers the exoskeleton movement energy through the collaboration of shape memory alloy springs and magnetorheological fluid brake devices. After being deformed by force, the shape memory alloy springs restore the preset shape due to temperature or stress changes, releasing the elastic deformation force, and converting mechanical energy into collectible energy; the magnetorheological fluid brake device adjusts the resistance in real time and cooperates with the spring to optimize the energy recovery process. The energy is stored in the 100mF supercapacitor group through DC-DC conversion, reducing energy waste, providing power for the exoskeleton system, extending the device's battery life, and reducing dependence on external power supplies.

[0062] For example, in the application scenario of intelligent prostheses, when the user goes up and down stairs, the shape memory alloy spring in the prosthesis senses the changes in joint force in real time. After being deformed by pressure, it relies on its own temperature or stress changes to restore the preset shape as the leg movement state changes, generating elastic deformation force to assist the flexion and extension of the knee joint, effectively alleviating the muscle burden of the user during exercise. It also cooperates with the magnetorheological fluid braking device to recover the excess energy generated by the movement, and stores it in the supercapacitor group through DC-DC conversion, realizing energy recycling. At the same time, it also makes the prosthetic movement smoother and more natural, improving the user's mobility and quality of life.

[0063] In step S105, based on the stable power supply of the supercapacitor group, the adaptive wheelbase adjustment mechanism of the multi-dimensional woven carbon fiber frame is used to match the wearer's leg length, and combined with the integrated joint servo mechanism, the power parameters are optimized, and the optimized power parameters are converted into joint motion torque. At the same time, the pressure distribution of the contact surface is dynamically adjusted through the pressure sensor array to perform human-machine collaborative movement and reduce the error rate.

[0064] Among them, the integrated joint servo mechanism integrates the servo motor, transmission mechanism, sensor and control system into the joint, and realizes high-precision position, speed and torque adjustment of joint movement through precise feedback control, providing efficient power and flexible movement for robotic arms, exoskeletons, etc.

[0065] It can be understood that the embodiments of the present application highly integrate the servo motor, transmission mechanism, sensor and control system into the joint, and use precise feedback control to achieve high-precision position, speed and torque adjustment of the joint movement, provide stable power output for robotic arms, exoskeletons and other equipment, optimize power parameters, determine normal operation, perform human-machine collaborative movement, dynamically adapt to the movement requirements under different working conditions, and reduce the movement error rate.

[0066] According to the embodiment of the present application, an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control is proposed. Through the flexible torque coupling mechanism with the help of planetary-magnetoelectric-fluid body and harmonic deceleration integrated structure, it can be continuously adjusted in the torque range of 0-600N·m and the stiffness range of 5-50N·m / rad, taking into account the power output of heavy-load operations and the flexibility requirements of fine movements, and adapting to complex motion scenes; holographic intelligent brain collaborative control uses quantum federated learning and bionic neural control algorithms to deeply integrate multi-source perception information, accurately analyze the user's motion intention and environmental data, and then dynamically optimize the torque strategy to make the exoskeleton The skeleton is highly coordinated with human movement, providing intelligent and efficient motion assistance even in complex scenarios. The integrated energy-safety shield module utilizes magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to create an energy recycling system and dual safety guarantees. This not only recycles motion energy to improve battery life, but also reduces impact risk through potential energy buffering. The ultra-lightweight integrated phantom structure utilizes multi-dimensional composite carbon fiber and a highly integrated design to achieve lightweight exoskeleton joints (weight reduction of over 30%) and high-speed response (joint movement delay <5ms), adapting to users of different body types and enhancing the comfort and flexibility of human-machine collaborative motion. This solves the problems of high energy consumption, complex mechanical structure, and insufficient stability in existing technologies.

[0067] The following will describe the exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control through a specific embodiment. Figure 11 Shown, including: Once the wearer puts on the exoskeleton, it's equipped with multiple sensors to collect data in real time. High-precision miniature six-axis inertial sensors (MPU6050) with ±0.1° accuracy are installed at each joint to monitor joint angle, angular velocity, and acceleration. A 4×4 pressure sensor array (FlexiForce A401) is placed on the soles of the feet to sense contact force between the body and the ground. High-precision cargo weight sensors are installed on the shoulder and back trays, communicating with the control module in real time via the CAN bus to obtain load weight data. A lidar (Velodyne VLP-16) and a binocular camera (Logitech C922) are integrated into the helmet to create a 360-degree environmental perception system, identifying terrain, obstacles, and other environmental information. All sensor data is transmitted at 100Hz to an edge computing platform powered by the NVIDIA Jetson AGX Orin for initial processing.

[0068] When the system receives motion scene data, the STM32H750 chip serves as the control core, collaboratively controlling the planetary gearset, magneto-electric drive unit, and rheological damper. For example, in a walking scenario, if the load is less than 20 kg, the STM32H750 chip reduces the current applied by the magneto-electric drive unit to the electromagnetic coil of the MR fluid damper, reducing the viscosity of the MR fluid and adjusting the joint stiffness to 5 N·m / rad. At this point, the planetary gearset operates at a lower transmission ratio, and the servo motor (such as the Maxon EC motor) operates at a relatively high speed to ensure joint flexibility. In a load-bearing scenario, if the load exceeds 20 kg, the chip increases the current, rapidly solidifying the MR fluid and increasing the stiffness to 50 N·m / rad. It also adjusts the planetary gearset ratio to amplify the motor output torque. Combined with the integrated implicit harmonic reduction structure, these chips generate the basic dynamic parameters required to meet the load requirements.

[0069] The generated basic dynamic parameters are fed into an integrated quantum federated learning control algorithm developed based on the TensorFlow framework. Simultaneously, myoelectric sensors (such as the Delsys Trigno Wireless) attached to key muscle groups collect muscle electrical signals. These signals, along with real-time joint angle data from six-axis inertial sensors at the joints, are fed into a biodynamic model. The quantum federated learning algorithm exchanges data and trains the model between edge computing nodes (local computing units within the exoskeleton) and cloud servers. The biomimetic neural control algorithm utilizes a long short-term memory (LSTM) model to deeply analyze the input data. For example, when the system detects forward shift of the center of gravity, increased knee flexion angle, and enhanced muscle electrical signals, it predicts the user's intention to stand up. It then generates a torque compensation strategy, reverse-driving the rheological damper to adjust the stiffness change rate and simultaneously adjusting the output duty cycle of the magnetoelectric unit to provide additional torque support to the joint.

[0070] During movement, when the joint rotates, the piston in the magnetorheological fluid brake device installed at the joint's axis of rotation drives the joint, pushing the magnetorheological fluid back and forth in the magnetic field, cutting the magnetic flux lines and generating an induced current. Simultaneously, the nickel-titanium alloy shape memory springs at the knee and hip joints deform during flexion and extension, storing elastic potential energy. This induced current, after being processed by the rectifier circuit, passes through a DC-DC converter along with the electrical energy converted from the potential energy released by the shape memory springs and is stored in a 100mF supercapacitor bank. The DC-DC converter utilizes a high-efficiency synchronous buck topology, achieving a conversion efficiency exceeding 95%, ensuring efficient energy recovery and storage.

[0071] The exoskeleton's multi-dimensional woven carbon fiber frame's adaptive wheelbase adjustment mechanism, powered by a stable power supply from a supercapacitor bank, operates. This mechanism, driven by a stepper motor and a screw-nut assembly, uses sensors to detect the wearer's leg length, completing wheelbase adjustment within five minutes to accommodate varying body types. An integrated joint servo mechanism (utilizing Kollmorgen's AKM series flat, frameless torque motors paired with 17-bit precision hollow encoders) converts electrical energy into precise joint motion torque based on optimized dynamic parameters. Simultaneously, an array of pressure sensors distributed throughout the exoskeleton's contact areas monitors pressure distribution across the contact surface in real time. When uneven pressure is detected, the system dynamically optimizes pressure distribution by adjusting joint angles and torque output, enabling collaborative human-machine motion. Field testing has shown that this process reduces motion error to less than 3%.

[0072] In summary, the embodiments of this application significantly improve operational efficiency and the wearable experience through precise multi-source data sensing, intelligent collaborative control, and efficient energy management. Multiple sensors collect motion, load, and environmental data at a high frequency of 100Hz, rapidly processed through edge computing, and provide real-time information for driving decisions. The STM32H750 chip intelligently controls the transmission unit, enabling adaptive switching between light-load flexibility (5N·m / rad stiffness) and heavy-load stability (50N·m / rad stiffness). Quantum federated learning, combined with biodynamic models, predicts motion intent and compensates torque within 0.5 seconds, resulting in more precise operational responses. Magnetorheological fluid and shape memory alloys collaborate to recover energy with over 95% efficiency, while a 100mF supercapacitor ensures endurance. Adaptive wheelbase adjustment provides personalized adaptation within 5 minutes, and pressure sensors dynamically optimize contact pressure, reducing the error rate in human-machine collaboration to less than 3%. This method effectively reduces labor intensity, improves efficiency and safety in scenarios such as logistics and rehabilitation, and promotes the practical development of exoskeleton technology.

[0073] Figure 12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include: A memory 1201 , a processor 1202 , and a computer program stored in the memory 1201 and executable on the processor 1202 .

[0074] When the processor 1202 executes the program, the exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control provided in the above embodiment is implemented.

[0075] Furthermore, the electronic device further includes: The communication interface 1203 is used for communication between the memory 1201 and the processor 1202 .

[0076] The memory 1201 is used to store computer programs that can be run on the processor 1202 .

[0077] The memory 1201 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0078] If the memory 1201, processor 1202, and communication interface 1203 are implemented independently, the communication interface 1203, memory 1201, and processor 1202 can be connected to each other via a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 12 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0079] Optionally, in a specific implementation, if the memory 1201, the processor 1202 and the communication interface 1203 are integrated on a chip, the memory 1201, the processor 1202 and the communication interface 1203 can communicate with each other through an internal interface.

[0080] The processor 1202 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0081] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control.

[0082] In addition, an embodiment of the present application also provides a computer program product, including a computer program or instructions. When the computer program or instructions are executed, the above-mentioned exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control is implemented.

[0083] In the description of this specification, reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0084] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0085] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0086] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logical functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0087] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above-mentioned embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0088] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. An exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control, characterized in that: include: Smart torque coupling mechanism, holographic intelligent brain collaborative control, energy-safety integrated shield module, ultra-lightweight integrated phantom structure; among them, The flexible torque coupling mechanism is designed to adapt to the power and flexibility of different complex motion scenarios. Through the coordination of the planetary-magnetoelectric-fluidic integrated transmission unit and the internal implicit harmonic reduction integrated structure, it can achieve 0-600N·m torque output and 5-50N·m / rad continuous adjustment of stiffness. The holographic intelligent brain collaborative control strategy is used to integrate sensory information, provide intelligent motion assistance in complex motion scenarios, combine quantum federated learning and bionic neural control algorithms, deeply analyze motion intentions and environmental data, optimize torque strategies and perform dynamic collaborative control; The energy-safety integrated shield module is used to construct a dual safety protection system, which uses magnetorheological fluid kinetic energy recovery and shape memory alloy potential energy storage to achieve energy recycling. The ultra-lightweight integrated phantom structure is used to adapt to different human body sizes and enhance human-machine collaborative movement. Through multi-dimensional composite carbon fiber materials and highly integrated design, lightweight adjustment and high-speed response of the exoskeleton joints are achieved.

2. The exoskeleton drive system with multimodal torque coupling and intelligent brain collaborative control according to claim 1, characterized in that: The dynamic torque coupling mechanism includes a planetary-magnetoelectric-fluidic integrated transmission unit and an implicit harmonic deceleration integrated structure. The planetary-magnetoelectric-fluidic integrated transmission unit is used to integrate planetary gear rigid transmission, magnetic powder clutch dynamic torque coupling and magnetorheological fluid flexible stiffness adjustment technology, and through the coordinated control of the three-stage planetary gear set and the magnetorheological fluid damping unit, a wide-range torque output and continuous adaptive adjustment of stiffness are performed; the implicit harmonic deceleration integrated structure is used to embed the ultra-precision harmonic reducer into the motor stator, combined with the collaborative design of the shape memory alloy wire-wound flexible wheel, to perform dynamic coupling control of stiffness.

3. The exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control according to claim 1, characterized in that: The holographic intelligent brain collaborative control includes a dual-loop intelligent adjustment system and a bionic neural-myoelectric fusion control algorithm. The dual-loop intelligent adjustment system is used to integrate the current-speed dual closed-loop control architecture and the quantum federated learning dynamic decision-making mechanism, and performs gait recognition, stiffness adjustment and power distribution through inner-loop real-time torque compensation and outer-loop energy-safety collaborative optimization; the bionic neural-myoelectric fusion control algorithm is used to integrate biological electromyographic signal real-time decoding and bionic central pattern generator dynamic modeling technology, and construct a three-level mapping model of electromyographic signal-neural intention-motion command through deep coupling of the surface electromyography sensor array and the LSTM neural network to perform real-time compensation of exoskeleton joint torque.

4. The exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control according to claim 3, characterized in that: The formula of the quantum federated learning dynamic decision-making mechanism is: in, is the updated global model parameter; is the global model parameter before updating; is the learning rate; N is the total number of nodes participating in federated learning; i, j are index variables; λ is a hyperparameter; is the Jensen-Shannon divergence; is the data distribution of the jth node; is the global data distribution; is an exponential function; is an operator for finding the gradient of model parameters; is the quantum state associated with the model parameters; is the quantum Hamiltonian; is the expectation value of the quantum Hamiltonian in the quantum state.

5. The exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control according to claim 1, characterized in that: The energy-safety integrated shield module includes a magnetic flow-elastic dual-effect energy management module and a Starlink encryption-neural reasoning security module. The magnetic flow-elastic dual-effect energy management module is used to convert motion kinetic energy into electrical energy in real time, and release elastic potential energy to assist joint movement during the gait swing phase; the Starlink encryption-neural reasoning security module is used to perform nanosecond-level secure evidence storage and real-time verification of control instructions, combining the 1D-CNN+LSTM neural network self-evolution to learn complex fault modes, and matching it with an independent hardware security chip for millisecond-level braking response, to build a triple security protection system of encryption evidence-intelligent diagnosis-hardware redundancy.

6. The exoskeleton drive system with multi-modal torque coupling and intelligent brain collaborative control according to claim 1, characterized in that: The ultra-lightweight integrated phantom structure includes a multi-dimensional composite carbon fiber lightweight module and a high-speed integrated execution module. The multi-dimensional composite carbon fiber lightweight module is used to dynamically adjust the axial stiffness and vibration attenuation during high-speed movement. It adopts a three-dimensional woven carbon fiber composite material and is combined with an adjustable joint mechanism formed by 3D printing. With an ultra-lightweight design of a single joint mass of less than 1.3 kg, it improves movement flexibility and adapts to the human body shape. The high-speed integrated execution module is used for nanosecond-level synchronous acquisition of sensor data, millisecond-level resolution of torque instructions, and sub-millisecond-level precise driving of magnetorheological fluid / magnetic powder clutch. It integrates a 2.5Mbps CANFD bus high-speed data link, a 22ms supercomputing edge node, and a 25kHz ultra-high frequency PWM drive technology to provide instant power response for high-speed movement of humans and machines.

7. An exoskeleton driving method with multimodal torque coupling and intelligent brain collaborative control applied to any one of claims 1-6, characterized in that: The method comprises: Obtain motion scene data; Based on the motion scene data, the joint stiffness is adjusted through the coordinated control of the planetary gear set, the magnetoelectric drive unit and the rheological damper to generate basic dynamic parameters suitable for scenes such as walking, running, jumping and weight bearing; The basic dynamic parameters are input into the control algorithm of integrated quantum federated learning. Combined with the biodynamic model, the real-time mapping of human joint angles and muscle electrical signals, the intelligent brain analyzes the movement intention and generates a torque compensation strategy, which reversely drives the rheological damper to adjust the stiffness change rate and the output duty cycle of the magnetoelectric unit. According to the torque compensation strategy, the kinetic energy is recovered through the magnetorheological fluid brake device and the shape memory alloy spring, and stored in the 100mF supercapacitor bank after DC-DC conversion; Based on the stable power supply of the supercapacitor group, the adaptive wheelbase adjustment mechanism of the multi-dimensional woven carbon fiber frame is used to match the wearer's leg length. Combined with the integrated joint servo mechanism, power parameter optimization is performed and the optimized power parameters are converted into joint motion torque. At the same time, the pressure distribution of the contact surface is dynamically adjusted through the pressure sensor array to perform human-machine collaborative movement and reduce the error rate.

8. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement an exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control according to claim 7.

9. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instructions are executed, the exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control of claim 7 is implemented.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instructions are executed, the exoskeleton driving method with multi-modal torque coupling and intelligent brain collaborative control of claim 7 is implemented.

Images