Simultaneous sensing method of touch and shape of flexible fingers based on local micro-deformation

Abstract

The present invention proposes a method for simultaneously sensing the touch and shape of a flexible finger based on local micro-deformation, which belongs to the field of perception of soft robots, including: setting a distributed curvature sensor according to the length of the finger to collect the local micro-deformation of the finger; through the distributed curvature sensor sensing data, the real-time tactile perception of the flexible finger in static and dynamic processes can be realized, including contact events, contact positions, and contact forces; through the same sensor, the body shape perception can also be realized in real time during the process of realizing tactile perception; based on the collected local micro-deformation, the local segment curvature is calculated to perceive the shape of the sensing finger. With the help of a sensor seamlessly integrated with the flexible finger, the present invention can perceive the micro-deformation of the corresponding area of ​​the sensing unit in real time; through simple numerical calculations, the real-time perception of the flexible finger's real-time shape, contact events, contact positions, and contact forces can be realized.

Description

Technical Field

[0001] The present invention belongs to the field of soft robot perception, and in particular relates to a method for simultaneously sensing the touch and shape of flexible fingers based on local micro-deformation. Background Art

[0002] Similar to body perception in natural systems, proprioceptive and tactile perception of soft robots plays a vital role in performing skilled tasks, perceiving their own shapes, perceiving the surrounding environment, and interacting with the surrounding environment. In recent years, people have conducted extensive research on proprioceptive and tactile perception of soft robots, exploring various methods, including fiber Bragg gratings, stretchable capacitive sensors, piezoelectric sensors, resistive strain sensors, stretchable optical waveguide sensors, magnetic sensors, etc. Despite the progress made in the perception of soft continuum robots, existing work still faces some problems and challenges. For example, proprioceptive and tactile perception need to rely on different sensing methods. Most proprioceptive shape perception is achieved through optical fiber sensing, while tactile perception relies on array-type capacitive or resistive sensors; this leads to problems such as complex structure of sensing components, poor real-time data acquisition and processing, difficulty in device integration, and complex implementation methods; in addition, the tactile perception of flexible fingers based on array sensors can only perceive the local force applied to the sensor, and cannot evaluate the impact of the applied force on the entire flexible finger.

[0003] Chinese invention patent CN14131580B discloses a drivable soft robot and its preparation method, shape sensing method and system. The Rayleigh backscattering band in the optical signal propagation process in the optical fiber is received through N single-mode sensing optical fibers in the drivable hose. The spatial morphology of the soft robot can be sensed, but the shape sensing device is complex and the economic cost is high.

[0004] Chinese invention patent CN114179119B discloses a flexible finger mechanism and soft hand with tactile and bending sensing functions. Using multiple array piezoresistive sensors, the contact force between the finger and the grasped object can be measured through piezoresistive sensors at different positions. Installing multiple bending sensors, the bending degree of different parts of the finger can be monitored through different bending sensors. The measurement of contact force and bending degree depends on different sensors, which makes the device complex, increases the integration constraints of the flexible finger, reduces reliability, and makes data processing difficult.

[0005] Chinese invention patent CN114770490B discloses a soft manipulator based on a cross-modal tactile sensor and a bubble driver, which relies on a complex visual system installed on the fingertips to collect the deformation of the columnar protrusions inside the soft film, complete the transformation from tactile mode to visual mode, and realize multifunctional perception. The force perception can only perceive the contact force of the fingertips. If the contact occurs in other parts of the finger, the force perception fails; the object geometry information is the geometry information of the contact surface between the grasped object and the fingertip visual sensor system. Due to the lack of perception of the shape of the finger body, the real grasping size cannot be obtained.

[0006] Chinese invention patent CN107139207B discloses a pneumatic flexible finger, a flexible finger control system and a control method, which realizes the measurement of the bending curvature and contact force of the finger by detecting the air pressure of the sensing air cavity. This method has some problems and shortcomings: 1) This method has special requirements for the structural design of the flexible finger, and a sensing air cavity needs to be specially set up, which has poor versatility; 2) This method senses the air pressure of the air cavity through air pressure, and indirectly predicts the curvature of the sensing finger after complex calculations. Indirect measurement and complex calculations will introduce large errors; 3) The curvature measurement conversion module and the predicted curvature calculation module assume that "the entire finger bending process conforms to the uniform curvature bending". In actual situations, due to factors such as self-gravity, external contact force, and material inhomogeneity, the flexible finger bending process does not conform to the overall uniform curvature bending (conforms to the segmented bending curvature); 4) When the method implements the contact force estimation, it assumes that "the length and width of the sensing air cavity bottom layer do not change when it is compressed, only the height changes", and based on this assumption, "the sensing air cavity bottom layer is simplified to a simply supported beam" to perform force analysis on it. However, the actual situation is similar to that of a flexible finger of a human being, which has only one fixed end. During use, only the root is fixed, and the part of the finger below the root can move freely. However, simplifying the force analysis of the finger into a "simply supported beam" model is quite different from the actual situation, and the analysis results obtained are bound to have a large deviation from the actual situation. Summary of the invention

[0007] To solve the above technical problems, the present invention provides a simultaneous sensing method for the touch and shape of flexible fingers based on local micro-deformations, wherein a distributed curvature sensor is arranged according to the length of the finger to collect the local micro-deformations of the finger; only one sensor is needed to simultaneously realize the real-time touch and shape perception of the flexible finger; the data sensed by the distributed curvature sensor can realize the real-time tactile perception of the flexible finger in static and dynamic processes, including contact events, contact positions, and contact forces. Different from the traditional array sensor that directly measures the contact force, the contact force sensing method of the present invention uses the cantilever beam equivalent model force analysis to calculate the equivalent contact force; through the same sensor, the body shape perception can also be realized in real time during the process of realizing tactile perception; based on the collected local micro-deformations, the local segment curvature is calculated to perceive the sensing finger shape.

[0008] The present invention can simultaneously realize real-time tactile perception and body shape perception of flexible fingers through a sensing means, realize flexible finger contact event perception, contact position perception, contact force perception, body shape perception and shape reconstruction, and combine tactile perception and shape perception to systematically evaluate the physical properties of the grasped object such as size and hardness.

[0009] In order to achieve the above object, the present invention provides the following technical solutions:

[0010] The method for simultaneously sensing the touch and shape of a flexible finger based on local micro-deformation uses a distributed curvature sensor to simultaneously realize the real-time touch perception, shape perception and shape reconstruction of the flexible finger, including the following steps:

[0011] S1000: setting a distributed curvature sensor according to the length of the flexible finger to collect local micro deformation of the flexible finger; the distributed curvature sensor includes a plurality of sensing units, and the sensing units are set starting from the root of the flexible finger;

[0012] S2000: Equivalent the flexible finger to a cantilever beam equivalent structure, and establish a flexible finger tactile perception model; realize real-time tactile perception of the flexible finger in static and dynamic processes according to the sensing data of the distributed curvature sensor, and the tactile perception includes contact events, contact positions, and contact forces; judge the contact events by the inductance differential of the sensing unit arranged at the root of the flexible finger, judge the contact position by the inductance differential of each sensing unit, and estimate the contact force by the contact position and the inductance change of the sensing unit arranged at the root of the flexible finger; the inductance differential is the differential of the inductance of the sensing unit with respect to time;

[0013] S3000: Utilizes distributed curvature sensors to achieve real-time shape perception and shape reconstruction of flexible fingers while realizing tactile perception; divides the flexible finger into segments, calculates the segment curvature based on the collected local micro-deformation, perceives the shape of the flexible finger, and realizes shape perception of the flexible finger in real time; takes the root fixed point of the flexible finger as the starting point, and the center of the first arc of the flexible finger as the coordinate center, establishes a shape reconstruction coordinate system, and based on the assumption that two adjacent arcs are tangent and the tangent point is collinear with the center of the two arcs, calculates the center, radius and starting point of each arc segment of the flexible finger, draws the segmented arc, and realizes shape reconstruction of the flexible finger.

[0014] The present invention has the following beneficial effects:

[0015] 1. The present invention can simultaneously realize tactile (contact event, contact position, contact force) and body shape perception through the same sensor, and reconstruct the body shape in real time, filling the current gap in the field of simple, compatible, shape and tactile integrated soft robot perception.

[0016] 2. The present invention directly measures the tiny deformation of the flexible finger body, and the sensing process is simple, reliable, rapid, highly accurate, and highly real-time. The force sensing method simplifies the flexible finger into a "cantilever beam" for force analysis, which conforms to the actual use characteristics of the finger and solves complex problems with a simple model.

[0017] 3. The present invention provides a simple, effective, universal and robust solution for soft robot perception, which can be seamlessly integrated on the surface of a flexible finger, move synchronously with the flexible finger, identify contact events, contact positions, estimate contact forces, perceive shape changes and reconstruct shape changes in real time. The present invention uses seamlessly integrated distributed sensors to directly measure tiny deformations. According to the movement characteristics of the finger, the finger is divided into multiple segments for contact force perception at different positions. The present invention realizes both tactile perception and body shape perception, and can obtain the geometric size, grasping force, tactile contact and sliding process of the grasped object during the grasping process. The present invention configures distributed sensors according to the bending characteristics of the flexible finger. The area corresponding to the sensing unit conforms to the equal curvature assumption, and the entire flexible finger conforms to the segmented equal curvature assumption. Such a division is closer to the actual situation, more accurately reflects the actual bending of the flexible finger, and has higher shape perception accuracy.

[0018] 4. According to the characteristic that one end of the flexible finger is fixed and the other end can move freely in actual application, the present invention simplifies the force analysis of the flexible finger into a "cantilever beam" model, which more accurately reflects the actual use of the flexible finger, and the obtained analysis result is more accurate. The present invention realizes the perception and reconstruction of the body shape through numerical calculation, and realizes the contact position perception and contact force estimation through the cantilever beam equivalent model. It does not require deep learning, complex model prediction, big data analysis and other means, and has good real-time perception, good reliability and high precision. It has low requirements on the hardware computing platform, which is conducive to reducing the power consumption of robot sorting, reducing the size and facilitating integration.

[0019] 5. In addition to realizing tactile perception and shape perception, the present invention can also obtain physical properties such as size and hardness of the grasped object.

[0020] 6. The method for real-time perception of flexible finger shape and touch proposed in the present invention is universal and can be used not only for flexible fingers, but also for shape detection, reconstruction, and tactile perception of other soft continuum robots.

[0021] In summary, the present invention realizes the shape perception and tactile perception of flexible fingers simultaneously through only one sensor, and has significant advantages such as simple structure, easy integration, low cost, and no influence on the movement of flexible fingers. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematic diagram of the simultaneous sensing method of touch and shape of flexible fingers based on local micro-deformation of the present invention;

[0023] Figure 2 Schematic diagram of the layout of distributed sensors on the surface of a flexible finger;

[0024] Figure 3 Schematic diagram of flexible fingers in different bending states;

[0025] Figure 4 A corresponding relationship diagram between the tactile perception contact area and the position of the sensor unit;

[0026] Figure 5 is a graph showing the relationship between the inductance of the sensing unit and the air pressure driving the flexible finger;

[0027] Figure 6 It is a schematic diagram of the real-time inductance change driving air pressure during multi-cycle cyclic bending;

[0028] Figure 7 Schematic diagram of the equivalent model of the cantilever beam based on contact force sensing of small deformation;

[0029] Figure 8 Schematic diagram of the change in inductance of the sensing unit when external contact is applied to different contact areas;

[0030] Fig. 9 Schematic diagram of the inductance differentiation of the sensing unit when external contact is applied for different contact areas;

[0031] Fig.10 Schematic diagram of differential analysis of the inductance of the sensing unit when an external contact is applied to the third contact area;

[0032] Fig.11 is the inductance change of the first sensing unit at the root when the contact occurs in the third contact area Schematic diagram of contact force relationship;

[0033] Fig.12 is the inductance change of the first sensing unit at the root when the contact occurs in the fourth contact area Schematic diagram of the relationship with contact force;

[0034] Fig.13 is the force perception coefficient k of the third contact area and the fourth contact area A4 F Schematic diagram of the relationship with driving air pressure changes;

[0035] Fig.14 A schematic diagram showing the tactile perception of contact force in different contact areas;

[0036] Fig.15 This is a comparison chart of the contact force tactile perception results and the commercial sensor detection results;

[0037] Fig.16 Schematic diagram of the corresponding relationship between the inductance change and curvature of different sensing units;

[0038] Fig.17 Schematic diagram of shape reconstruction method based on distributed micro-deformation monitoring;

[0039] Fig.18 This is a shape comparison diagram of real-time shape reconstruction and visual detection of flexible fingers in different bending states;

[0040] Fig.19 is a schematic diagram of a two-finger gripper;

[0041] Fig. 20 The real-time reconstruction result of the shape of the two-finger gripper when gripping the sponge along the length direction of the sponge (contact occurs in the fourth contact area);

[0042] Fig.21 A schematic diagram of the real-time distance between the two-finger gripper at the center of the fourth contact area when gripping the sponge along the length (10 cm) of the sponge;

[0043] Fig. 22 This is the real-time force sensing result diagram of the two fingers of the gripper during the process of grasping a fresh cake (36.9 g).

[0044] Among them, the figure markings are: distributed curvature sensor 1, columnar silicone segment 2, air cavity segment 3, first sensing unit S1, second sensing unit S2, third sensing unit S3, fourth sensing unit S4, first contact area A1, second contact area A2, third contact area A3, fourth contact area A4. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0046] The following will refer to Figures 1 to 22 Embodiments of the present invention are described in detail. Although specific embodiments of the present invention are shown in the accompanying drawings, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0047] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0048] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0049] In the case of using expressions such as "at least one of A, B, and C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.). In the case of using expressions such as "at least one of A, B, or C, etc.", it should generally be interpreted in accordance with the meaning of the expression generally understood by those skilled in the art (for example, "a system having at least one of A, B, or C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0050] To facilitate understanding of the embodiments of the present invention, further explanation will be given below by taking specific embodiments as examples in conjunction with the accompanying drawings, and the various drawings do not constitute a limitation on the embodiments of the present invention.

[0051] In an implementation case, Figure 1 As shown, the present invention provides a method for simultaneously sensing the touch and shape of a flexible finger based on local micro-deformation, comprising the following steps:

[0052] S1000: Distributed curvature sensor 1 is set according to the length of the finger to directly collect the local micro-deformation of the flexible finger; only one sensor is needed to realize the real-time perception of the touch and shape of the flexible finger at the same time. The distributed curvature sensor described in this embodiment is a distributed curvature sensor, which is composed of 4 curvature sensors based on crack gap magnetoresistance modulation. Each curvature sensor based on crack gap magnetoresistance modulation serves as a sensing unit. The length of the distributed curvature sensor 1 is 100mm, and the position of the distributed curvature sensor 1 is set from the root of the flexible finger. The flexible finger includes a columnar silicone segment 2 that does not participate in the movement and an air cavity segment 3 that produces bending deformation. The root refers to the position where the columnar silicone segment 2 and the air cavity segment 3 are connected.

[0053] The curvature sensor based on crack gap magnetoresistance modulation consists of a flexible PCB coil and a crack ferrite film. The flexible PCB coil is bonded to the upper surface of the crack ferrite film through a silicone adhesive, and the lower surface of the crack ferrite film is bonded to the surface of the flexible finger through a silicone adhesive.

[0054] like Figure 2 As shown, the distributed curvature sensor 1 is integrated on the surface of the flexible finger through a flexible adhesive, and there is no gap between the two. The soft distributed curvature sensor 1 follows the movement of the flexible finger without affecting the movement of the finger, and senses the local micro-deformation of the surface of the flexible finger in real time.

[0055] The flexible finger adopts the classic pneumatic network structure design sensing soft finger. The shell is printed with polylactic acid material, and the internal support structure is printed with water-soluble polyvinyl alcohol material. After the shell and internal support are assembled, the soft pneumatic finger is cast in one step with mixed and degassed silicone, and moves through silicone actuators.

[0056] like Figure 3 Schematic diagram of the soft finger under different actuation pressures is shown. Experiments show that the soft finger can bend over 180 degrees without being affected by the integrated distributed curvature sensor 1.

[0057] In S1000, Figure 4 As shown, this embodiment includes four sensing units, that is, the distributed curvature sensor 1 includes the first sensing unit S1 (0mm-25mm from the base of the flexible finger), the second sensing unit S2 (25mm-50mm from the base of the flexible finger), the third sensing unit S3 (50mm-75mm from the base of the flexible finger), and the fourth sensing unit S4 (75mm-100mm from the base of the flexible finger). The flexible finger provided with the distributed curvature sensor 1 is divided into 4 sensing areas, which correspond to the four sensing units one by one, namely the first contact area A1 (0mm-25mm from the base of the finger), the second contact area A2 (25mm-50mm from the base of the finger), the third contact area A3 (50mm-75mm from the base of the finger), and the fourth contact area A4 (75mm-100mm from the base of the finger).

[0058] S2000: Through the sensing data of the distributed curvature sensor 1, the real-time tactile perception of the flexible finger in static and dynamic processes can be realized, including contact events, contact positions, and contact forces. Different from the traditional array sensor that directly measures the contact force, the contact force sensing method of the present invention simplifies the force of the flexible finger into an equivalent model of a cantilever beam, and calculates the equivalent contact force based on its force analysis; when the flexible finger undergoes a slight deformation, the deformation amount will be detected by the distributed curvature sensor 1, and the slight deformation amount will be calculated based on the change in the inductance value of the first sensor unit S1, the second sensor unit S2, the third sensor unit S3, and the fourth sensor unit S4. Figure 5 As shown, during the bending of the flexible finger, the inductance of the sensing unit changes monotonically with the driving air pressure inside the flexible finger.

[0059] Inductance change is defined as follows: ;

[0060] in, is the inductance change, L t is the inductance of the coil at time t, and L0 is the inductance of the coil in the straight state of the distributed curvature sensor.

[0061] like Figure 6 As shown in the figure, the first sensing unit S1, the second sensing unit S2, the third sensing unit S3, and the fourth sensing unit S4 of the distributed curvature sensor 1 change the inductance as the driving air pressure of the flexible finger changes in three pressure increase and decrease cycles. It presents stable and repeatable regular changes. It has fast response, high sensitivity and no hysteresis. The experimental results show that the distributed curvature sensor 1 can detect the change of cyclic bending curvature with high resolution and good stability.

[0062] In another embodiment, the S2000 further includes the following steps:

[0063] S2100: The flexible finger is equivalent to a cantilever beam structure, and the tactile force perception model of the flexible finger is established as follows:

[0064] ;

[0065] Where F is the contact force, is the change in inductance of the first sensing unit after external contact occurs, It is calculated The force perception coefficient.

[0066] In this step, if Figure 7 As shown, in actual use, the flexible finger needs to be fixed at one end and the other end is free to bend. Generally, the columnar silicone segment 2 at the root of the flexible finger is fixed with a structural member. At this time, the flexible finger can be equivalent to a cantilever beam. The equivalent cantilever beam model is used to study the stress of the body of the flexible finger when an external load is applied.

[0067] When external loads (such as external forces) are applied to different contact areas of the flexible finger, the contact force The inductance change of the first sensing unit after the flexible finger is touched by an external force , resulting in a small deformation centered on the root fixed point, and the bending moment caused by this external force for:

[0068] ;

[0069] Where d is the vertical distance between the loading point and the fixed end of the flexible finger.

[0070] Further, S2100 includes the following steps:

[0071] S2101: Based on the sensing data of the straight state of the flexible finger, the change amount and change rate of the local micro deformation during the bending process of the flexible finger are calculated as data input for tactile perception. The local micro deformation refers to the micro deformation of the corresponding area of ​​each sensing unit. The bending deformation of the micro deformation is ≤ 0.01°.

[0072] In this step, if Figure 8 As shown, the bending moment M causes the inductance value of each sensing unit of the flexible finger to change. When external force is applied to the fourth contact area A4, the third contact area A3, and the second contact area A2 of the flexible finger, the change in the inductance value of the first sensing unit S1 at the root of the flexible finger is always greater than 0.

[0073] In this step, if Fig. 9 As shown, when different contact areas come into contact with the outside, the inductance differentials dL / dt of the four sensing units show different change rules, and the occurrence time and location of the contact event can be identified based on this change rule.

[0074] S2102: Contact events and contact positions are perceived based on the overall change rules of local micro-deformations at different positions during the contact process; when external contact occurs at different positions of the flexible finger, the change speeds of the micro-deformation data detected by each distributed sensor unit have different rules; the data change speed of the sensor unit at the force point is less than 0, the data change speed of the sensor unit between the force point and the root of the finger is greater than 0, and the data change speed of the sensor unit between the force point and the end of the flexible finger is less than 0; based on the data change speed of the sensor unit of the flexible finger, the contact event and contact position of the sensing finger can be determined.

[0075] In this step, further, Fig.10 As shown in the figure, the change law of the inductance change rate is analyzed by taking the external contact of the third contact area A3 as an example. In the loading stage (15s ~ 16.5s), the inductance differential dL / dt of the first sensor unit S1 and the second sensor unit S2 is positive, and the inductance differential dL / dt of the third sensor unit S3 and the fourth sensor unit S4 is negative. The dL / dt of all sensors reverses in the unloading stage (18s ~ 20s), and returns to zero after the unloading is completed and the soft finger reaches a stable state (i.e. Fig.10 High-frequency noise is observed on the inductance differential curve, which indicates that tiny vibrations occur when the external force is removed from the flexible finger.

[0076] In this step, the contact area / position can be identified by the value of the inductance differential dL / dt of the four sensors in the loading phase. The specific rules are shown in Table 1.

[0077] Table 1 Contact area identification rules

[0078]

[0079] S2103: Based on the cantilever beam equivalent model of the contact force of the flexible finger, the contact force at different contact positions is calculated through the slight deformation of the root of the flexible finger; a corresponding relationship between the contact force at different positions of the flexible finger and the slight deformation of the root of the finger is established; when the flexible finger is subjected to force at different positions, the root of the finger will undergo slight deformation, and the slight deformation causes the sensor data of the first sensor unit S1 at the root of the finger to change; at the same position, a corresponding relationship can be established between the magnitude of the force and the change in the sensor data (based on the sensor data of the first sensor unit S1 at the root at the moment of contact).

[0080] When the flexible finger grasps an object, the contact with the object mostly occurs in the third contact area A3 and the fourth contact area A4. The present invention elaborates on the contact force perception occurring in the third contact area A3 and the fourth contact area A4. Due to different contact position areas, the bending moment arm of the contact force on the root of the flexible finger is different, so the bending moment applied by the contact force at the root of the flexible finger in different contact position areas is also different; by applying contact force to different contact areas of the flexible finger in different bending states (different internal driving air pressures), the slight deformation of the root of the flexible finger caused by the contact force is detected, and the corresponding relationship between the contact force at different positions of the flexible finger and the slight deformation of the root of the flexible finger is established.

[0081] In this step, if Fig.11 , Fig.12 As shown, when the third contact area A3 and the fourth contact area A4 of the flexible finger are subjected to external contact force in different bending states of the flexible finger (different internal driving air pressure), the inductance increment of the corresponding sensing unit S1 at the root of the finger Contact force F under different air pressures and the sensor unit inductance increment of the first sensor unit S1 at the root of the finger For a specific sensing finger state and loading area, the slope k of these linear fitting lines is F (Force Perception Coefficient) is fixed.

[0082] In this step, if Fig.13 As shown, the force perception coefficient k F As the driving pressure increases, the force perception coefficient k of the contact force in the third contact area A3 under the same air pressure increases. F Greater than the force perception coefficient k of the contact force in the fourth contact area A4 F Therefore, once the contact position is identified, the sensor unit inductance increment of the first sensor unit S1 can be used And force perception coefficient k FCalculate the real-time contact force and estimate the contact force of the sensing finger.

[0083] Force perception coefficient k of the contact force in the third contact area A3 of the flexible finger under different bending states F The corresponding relationship with the driving air pressure p inside the flexible finger is as follows:

[0084] ;

[0085] Force perception coefficient k of the contact force in the fourth contact area A4 of the flexible finger under different bending states F The corresponding relationship with the driving air pressure p inside the flexible finger is as follows:

[0086] ;

[0087] S2104: When contact events occur at different positions, the size of the contact force acting on different contact areas has a corresponding relationship with the change in the sensor data of the sensor unit at the root of the finger. According to the change in the sensor data of the first sensor unit S1 at the root of the finger after the contact event occurs, the size of the contact force exerted on the flexible finger is estimated according to the flexible finger tactile force perception model.

[0088] In this step, an implementation example of the present invention is as follows: Fig.14 As shown, the contact position and contact force of the flexible finger pushed by the glass rod are sensed in real time. The sensing finger is pushed by the glass rod (internal driving pressure is 20 kPa); when t1=8 s, the external load is applied to the fourth contact area A4, and the peak force is about 0.2 N; when t2=17 s, the external load is applied to the third contact area A3, and the peak force is about 0.3 N; when t3=30 s, the external load is applied to the fourth contact area A4, and the peak force is about 1.0N.

[0089] In this step, an embodiment of the present invention, such as Fig.15 As shown, the contact force sensing results are compared with commercial sensors. The contact force estimation results are compared with the real force monitored by a commercial sensor (LSB200, Futek, USA); Fig.15 The range of the force intensity diagram is 0~0.6 N. A nylon rod is used to push the sensing finger, and a commercial force sensor for force comparison is installed on the nylon rod. When t1=21 s, an external load is applied to the third contact area A3, and the peak force is about 0.35 N. When t2=32 s, an external load is applied to the contact area A4, and the peak force is about 0.22 N. During the test, the estimated force is very consistent with the actual force, and the root mean square error RMS is less than 0.02 N.

[0090] S3000: Through the same sensor, the shape of the body can be perceived in real time during the process of tactile perception. The present invention directly measures the local micro-deformation of the flexible finger body, calculates the local segment curvature based on the collected local micro-deformation, and perceives the shape of the sensing finger.

[0091] In another embodiment, the S3000 further includes the following steps:

[0092] S3101: Based on the assumption that the shape change of the flexible finger during bending conforms to the segmented equal curvature model, the flexible finger is divided into several contact areas according to the bending characteristics of the flexible finger, so that during the bending of the flexible finger, the bending deformation in each small contact area conforms to the assumption of equal curvature. Sensing units are arranged in the corresponding contact areas. By sensing local micro-deformations, the segmented curvature of the finger can be obtained, thereby further realizing real-time shape perception of the finger.

[0093] S3102: Calibrate the bending state of the flexible finger and establish a corresponding relationship between the segmented curvature and the inductance change.

[0094] In this step, the “inductance change-curvature” relationship curves of different contact areas are obtained through the visual calibration method, such as Fig.16 shown.

[0095] In this step, the "inductance curvature sensing model" is obtained by numerically fitting the "inductance change-curvature" relationship curve. i and inductance change Linear correlation. By detecting the change in inductance of the sensing unit, the real-time tiny deformation of the segmented contact area of ​​the flexible finger can be sensed in real time.

[0096] ;

[0097] Wherein, K1 is the segmental curvature of the first contact area A1, is the inductance variation of the first sensing unit S1; K2 is the segmented curvature of the second contact area A2, K3 is the segmented curvature of the third contact area A3, is the inductance variation of the third sensing unit S3; K4 is the segmented curvature of the fourth contact area A4, is the inductance variation of the fourth sensing unit S4;

[0098] S3103: Based on the relationship between the segmented curvature and the inductance change, the real-time segmented curvature of the flexible finger is calculated according to the inductance change.

[0099] S3104: With the fixed point at the root of the finger as the starting point and the center of the first arc as the coordinate center, a shape reconstruction coordinate system is established, such as Fig.17 shown. Fig.17 In the figure, O1 is the arc center of the first contact area A1, θ1 is the arc angle of the first contact area A1, and r1 is the arc radius of the first contact area A1; O2 is the arc center of the second contact area A2, θ2 is the arc angle of the second contact area A2, and r2 is the arc radius of the second contact area A2; O3 is the arc center of the third contact area A3, θ3 is the arc angle of the third contact area A3, and r3 is the arc radius of the third contact area A3; O4 is the arc center of the fourth contact area A4, θ4 is the arc angle of the fourth contact area A4, and r4 is the arc radius of the fourth contact area A4.

[0100] S3105: Based on the assumption that two adjacent arcs are tangent and the tangent point is collinear with the centers of the two arcs, the center, radius and starting point of the segmented arc are calculated to achieve real-time perception and reconstruction of the flexible finger shape, such as Fig.17 shown.

[0101] In this step, the present invention combines tactile perception and shape perception to evaluate the physical properties of the grasped object, which include size and hardness. When analyzing the shape perception during the grasping process, a robot coordinate system is established with the midpoint of the line connecting the roots of the two fingers as the center of the circle. The spatial shape changes of the two flexible fingers during the grasping process can be obtained in real time through shape perception, and the spatial coordinates of the flexible fingers in the robot coordinate system can be further determined. When contact occurs, the contact position can be obtained through contact perception. Combined with the spatial coordinates of the flexible fingers, the spatial distance of the contact area of ​​the two flexible fingers can be obtained, which is the size of the grasped object, and the size attribute of the grasped object can be evaluated.

[0102] By recording the detected size of the grasped object at the beginning of contact during the grasping process and the size of the grasped object at the stable grasping stage, the size change from the beginning of contact to the stable grasping stage is obtained. Combined with the perceived contact force, the larger the ratio of size change to contact force (mm / N), the greater the object deformation caused by 1N force, indicating that the hardness of the grasped object is low, and the hardness of the grasped object can be evaluated.

[0103] like Fig.18 As shown, by changing the driving air pressure of the flexible finger to drive the flexible finger to bend freely, the shape of the flexible finger is reconstructed in real time and compared with the trial shape tracked by image vision. The higher the air pressure, the closer the result of shape perception reconstruction is to the visual tracking result.

[0104] An embodiment of the present invention, as Fig.19As shown in the figure, a two-finger (left finger and right finger) flexible sensing gripper was developed using flexible fingers. The gripper consists of two sensing fingers mounted on a 3D printed frame. The two fingers are connected together by a mechanical hinge structure. The hinge structure has an adjustment function, which can adjust the distance and opening degree of the fingers according to the size of the grasped object. The maximum distance is 80 mm and the maximum opening angle is 90°. Each flexible finger integrates a distributed curvature sensor (including four sensing units) and an inductance measurement circuit board designed based on the LDC1614 chip. The sensing data of the left and right fingers are measured independently, and the proprioception and tactile perception of the left and right fingers are also performed independently. The control of the left and right fingers can be performed independently or synchronously. The distributed curvature sensor can sense the shape of the left and right fingers in real time, estimate the external contact position and contact force, realize force control closed-loop control, perform controllable operations, and complete refined grasping tasks.

[0105] In one embodiment of the present invention, a cleaning sponge (5.43 g) is grasped along the height (4 cm), width (6 cm) and length (10 cm) of the cleaning sponge with a preset force of 0.1 N. By calculating the distance between the contact positions of the two fingers, the finger shape reconstruction result at the time of contact is obtained, which can provide information on the size of the object. The estimated target sizes are 3.6518 cm, 5.5175 cm and 9.7612 cm, respectively, which are consistent with the actual values. Fig. 20 The shape of the two-finger gripper is reconstructed in real time when the sponge is clamped along the length direction of the sponge. The contact occurs in the fourth contact area A4. Fig.21 When clamping the sponge along the length (10 cm) of the sponge, the real-time distance of the center position of the two-finger clamp contacting the fourth contact area A4 reflects the size information of the clamped object.

[0106] One embodiment of the present invention is the grasping of three different objects of similar weight but different hardness. In order to prevent damage to the delicate cap of fresh mushrooms (21.3 g), the grasping force is set to 0.1 N. The grasping forces of fresh cakes (36.9 g) and rigid Rubik's cubes (53 g) are set to 0.4 N and 1 N respectively. Fig. 22 This is the real-time force sensing result of the two fingers of the gripper during the process of grabbing the cake.

[0107] These demonstrations show that the flexible finger shape and tactile real-time perception amplification proposed by the present invention can achieve robust and accurate flexible finger tactile sensing under different contact conditions, and the tactile sensing method based on local deformation can accurately estimate the total force applied to the flexible finger. It brings new ideas for the application of robot tactile sensing and human-computer interaction.

[0108] The present invention mainly provides a simultaneous sensing method for the touch and shape of flexible fingers based on local micro-deformation. Based on the local micro-deformation of the flexible finger, a suitable sensor and sensor step size are selected, and the real-time shape perception and reconstruction of the flexible finger can be realized through a sensor, and the real-time contact position perception of the flexible finger and the estimation of the contact force can be realized through simple numerical calculation. The method provided by the present invention can also be applied to soft grippers, which can perceive the grasping process and grasping force information in real time and realize closed-loop force feedback control. The above description is only a relatively simple and effective implementation case of the present invention. The flexible finger shape and tactile perception method of the present invention has stable performance and system integration. It only needs to measure the inductance of the coil to realize the perception of the bending shape, contact position, and contact force, highlighting the versatility, high performance and easy implementation of the present invention. It will support the sensing ability of the touch of new soft robots, and open up many possibilities for its application in the fields of touch, VR / AR, robots, and medical treatment.

[0109] Parts of the present invention that are not disclosed in detail belong to the common knowledge in the art.

[0110] Although the above describes the illustrative specific embodiments of the present invention to facilitate those skilled in the art to understand the present invention, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations using the concept of the present invention are protected.

Claims

1. A method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation, characterized in that: Using a distributed curvature sensor to simultaneously realize real-time tactile perception, shape perception, and shape reconstruction of a flexible finger includes the following steps: S1000: setting a distributed curvature sensor according to the length of the flexible finger to collect local micro deformation of the flexible finger; the distributed curvature sensor includes a plurality of sensing units, and the sensing units are set starting from the root of the flexible finger; S2000: Equivalent the flexible finger to a cantilever beam equivalent structure, and establish a flexible finger tactile perception model; realize real-time tactile perception of the flexible finger in static and dynamic processes according to the sensing data of the distributed curvature sensor, and the tactile perception includes contact events, contact positions, and contact forces; determine the contact events by the inductance differential of the sensing unit arranged at the root of the flexible finger, determine the contact position by the inductance differential of each sensing unit, and estimate the contact force by the contact position and the inductance change of the sensing unit arranged at the root of the flexible finger; The inductance differential is the differential of the inductance of the sensing unit with respect to time, including: S2100: Equivalent the flexible finger to a cantilever beam equivalent structure, and establish the flexible finger tactile perception model as follows: ； in, is the contact force, is the change in inductance of the first sensing unit after external contact occurs, is to calculate the inductance change of the first sensing unit The force perception coefficient of the first sensing unit is a sensing unit arranged at the root of the flexible finger; S2101: Based on the sensing data of the straight state of the flexible finger, the change amount and data change speed of the local micro-deformation in the bending process of the flexible finger are calculated as data input for tactile perception; S2102: realizing the perception of the contact event and the contact position according to the data change speed of the local micro-deformation of different contact areas of the flexible finger during the contact process; the data change speed has the following rule: the data change speed of the sensing unit at the force point is less than 0, the data change speed of the sensing unit between the force point and the root of the flexible finger is greater than 0, and the data change speed of the sensing unit between the force point and the end of the flexible finger is less than 0; S2103: According to the cantilever beam equivalent structure of the flexible finger, by detecting the local micro-deformation of the root of the flexible finger, the contact force magnitudes of different contact areas are calculated according to the flexible finger tactile perception model; S3000: Utilizes distributed curvature sensors to achieve real-time shape perception and shape reconstruction of flexible fingers while realizing tactile perception; divides the flexible finger into segments, calculates the segment curvature based on the collected local micro-deformation, perceives the shape of the flexible finger, and realizes shape perception of the flexible finger in real time; takes the root fixed point of the flexible finger as the starting point, and the center of the first arc of the flexible finger as the coordinate center, establishes a shape reconstruction coordinate system, and based on the assumption that two adjacent arcs are tangent and the tangent point is collinear with the center of the two arcs, calculates the center, radius and starting point of each arc segment of the flexible finger, draws the segmented arc, and realizes shape reconstruction of the flexible finger.

2. The method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation according to claim 1, characterized in that: In S1000, the number and distribution positions of the sensing units are determined by the length of the flexible finger and the sensing accuracy requirements.

3. The method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation according to claim 1, characterized in that: The S3000 includes: The shape change of the flexible finger during the bending process conforms to the piecewise equal curvature hypothesis. The contact area is set, and the piecewise curvature of the flexible finger is obtained by sensing the local slight deformation of the flexible finger, thereby realizing real-time shape perception of the finger.

4. The method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation according to claim 3 is characterized in that: The S3000 further includes: The bending state of the flexible finger is calibrated, and the corresponding relationship between the segmented curvature and the inductance change is established.

5. The method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation according to claim 4, characterized in that: The S3000 further includes: Based on the corresponding relationship between the segmented curvature and the inductance variation, the real-time segmented curvature of the flexible finger is calculated according to the inductance variation.

6. The method for simultaneously sensing touch and shape of flexible fingers based on local micro-deformation according to claim 5, characterized in that: The physical properties of the grasped object, including size and hardness, are evaluated by combining tactile perception and shape perception.

Images