A control method for Bowden wire and binding fabric drive

Abstract

The present invention provides a control method for Bowden wire and binding fabric transmission, comprising the following steps: The processor uses a hybrid tension compensation model to obtain a hybrid tension compensation value; The expected tension value and the hybrid tension compensation value are summed to calculate the final tension input value; Using the final tension input value as the actual input of the system, a fuzzy auto-disturbance rejection controller is adopted to obtain an input tension control signal for the drive motor, thereby controlling the drive motor. The drive motor acts on the human leg through a steel wire rope, so that the actual tension received by the leg is the expected tension value. A control method for Bowden wire and binding fabric transmission provided by the present invention has the following advantages: The control method for Bowden wire and binding fabric transmission provided by the present invention can effectively improve the control accuracy of Bowden wire and binding fabric during transmission.

Description

Technical Field

[0001] The present invention belongs to the technical field of service robots, and particularly relates to a control method for Bowden wire and binding fabric transmission. Background Art

[0002] In recent years, the remarkable functions of flexible lower limb assistive exoskeletons in the fields of medical treatment, handling, and military have attracted extensive attention from researchers. Motor drive technology is currently relatively mature and easy to control, and it is a commonly used drive method for flexible lower limb assistive exoskeletons. The method of driving the Bowden wire from the motor output end to the binding fabric on the human leg is the most common flexible assistive device for flexible exoskeletons.

[0003] At present, the force transmission path of flexible lower limb assistive exoskeletons: Bowden wire - binding fabric has problems such as force loss, hysteresis, and poor control accuracy during the transmission process. For the Bowden wire force transmission system, the patent "A Flexible Knee Joint Assistive Rehabilitation Device Based on Lasso Transmission" designs a structure that transmits the force from the motor output through the lasso to the end nylon binding fabric, achieving flexibility in structure and transmission. The device of this patent is simple and lightweight, focusing on the structural design, but less involved in the control of this transmission system. The literature "Analysis of Lasso Transmission Characteristics" analyzes the static model and dynamic model of the lasso force and displacement transmission characteristics by building an experimental platform, and obtains that the non-linear characteristics such as dead zone, hysteresis, and direction dependence in the input-output relationship of lasso force and displacement are related to the friction coefficient between the lasso contact area, the total curvature of the lasso, and the equivalent elastic coefficient of the rope. The literature "Analysis of Torque Transmission Characteristics and Friction Compensation of Double Lasso System" deduces the torque transmission model of double lasso under any load conditions, and designs a friction compensation controller without end feedback for the torque transmission model, so that the output of the system can follow the desired torque trajectory. The above literature studies the influencing factors and torque transmission characteristics during the lasso transmission process, and makes friction compensation, but lacks the force transmission analysis of the end binding fabric. Summary of the Invention

[0004] Aiming at the defects existing in the prior art, the present invention provides a control method for Bowden wire and binding fabric transmission, which can effectively solve the above problems.

[0005] The technical solution adopted by the present invention is as follows:

[0006] The present invention provides a control method for Bowden wire and binding fabric transmission, including the following steps:

[0007] Step 1, construct a flexible lower limb assistive exoskeleton; the flexible lower limb assistive exoskeleton includes a processor, a drive motor, a Bowden wire, a binding fabric, and a perception and acquisition unit;

[0008] The Bowden cable includes a steel wire rope and a casing sleeved outside the steel wire rope; one end of the steel wire rope is connected to the output end of the driving motor; the other end of the steel wire rope is connected to the binding fabric; the binding fabric is fixed on the human leg;

[0009] Step 2, when a person wears the flexible lower limb assistive exoskeleton to walk, the perception and acquisition unit real-time perceives and obtains the human leg movement information, and sends the human leg movement information to the processor;

[0010] Step 3, the processor analyzes the human leg movement information, obtains the current gait cycle, and thus generates an expected pulling force value F0;

[0011] Step 4, the processor uses a hybrid pulling force compensation model to obtain a hybrid pulling force compensation value F3;

[0012] Step 4.1, perform a micro-element analysis on the Bowden cable drive, establish a static friction model of the Bowden cable, and obtain a friction compensation value F1 of the Bowden cable at the current moment according to the static friction model of the Bowden cable;

[0013] Step 4.2, perform a numerical simulation analysis of the tensile mechanical properties of the binding fabric by using the finite element method, and establish a tensile force compensation model of the binding fabric; obtain a tensile force compensation value F2 of the binding fabric at the current moment according to the tensile force compensation model of the binding fabric;

[0014] Step 4.3, perform a summation operation on the friction compensation value F1 of the Bowden cable and the tensile force compensation value F2 of the binding fabric to obtain a hybrid tensile force compensation value F3;

[0015] Step 5, perform a summation calculation on the expected pulling force value F0 and the hybrid pulling force compensation value F3 to obtain a final pulling force input value F4;

[0016] Taking the final pulling force input value F4 as the actual input of the system, using a fuzzy auto-disturbance rejection controller, obtain an input pulling force control signal for the driving motor, so as to control the driving motor. The driving motor acts on the human leg through the steel wire rope, and makes the actual pulling force received by the leg equal to the expected pulling force value F 0。

[0017] Preferably, step 4.1 is specifically:

[0018] The static friction model of the Bowden cable describes the pulling force transmission relationship between the input end and the output end of the Bowden cable. Specifically, when the Bowden cable is stationary, F out =F in (t - ); where, F out is the pulling force at the output end of the Bowden cable at the current moment; F in (t -) is the tension at the Bowden cable input end at the previous moment;

[0019] When the Bowden cable reciprocates, where, is the relative sliding speed of the wire rope inside the casing, and is specified as the positive direction when it is consistent with the input tension direction, and negative otherwise; μ is the friction coefficient between the wire rope and the casing, κ(l,t) is the curvature at this place, F out (t - ) is the output tension at the previous moment; F in is the tension at the Bowden cable input end at the current moment; l is the length of the infinitesimal segment of the Bowden cable; t is the current time; L is the total length of the Bowden cable.

[0020] Preferably, step 4.2 is specifically:

[0021] Establish a lower limb and binding fabric model; the lower limb and binding fabric model includes the thigh, calf, leg binding fabric and end binding fabric; a hinge connection is provided at the connection between the thigh and the calf; the leg binding fabric is externally bound to the calf; the upper part of the leg binding fabric is the end binding fabric;

[0022] After dividing the meshes of each component in the lower limb and binding fabric model, use the surface-to-surface binding method to define the interaction between the calf and the leg binding fabric, and the leg binding fabric is connected to the end binding fabric by the binding method;

[0023] Then apply a full fixation constraint to the thigh, and apply a gravity load to all the remaining components to simulate the gravity received during the walking process;

[0024] Then apply a pressure load in the positive Y direction above the end binding fabric, and the magnitude of the load is consistent with the joint moment required by the person during the gait cycle. Therefore, the applied load magnitude is a sine curve with both amplitudes being positive, and the periodic amplitude loading is selected to ensure that the calf reciprocates within the period;

[0025] By outputting the history variables at intervals of 1 / 4 on the upper surfaces of the leg binding fabric and the end binding fabric respectively, and then post-processing the history variable output results of the sectional resultant force, the change of the output force received by each section is shown in the post-processed curve;

[0026] Plot the relationship between the load magnitude input from the input end of the binding fabric and the overall average output tension received by the end binding fabric, obtain the relationship between the input force and the output force of the binding fabric, and further obtain the force transfer model of the binding fabric;

[0027] According to the force transfer model of the binding fabric, obtain the tension compensation value F2 of the binding fabric at the current moment.

[0028] The control method for Bowden cable and binding fabric transmission provided by the present invention has the following advantages:

[0029] The control method for Bowden cable and binding fabric transmission provided by the present invention can effectively improve the control accuracy of the Bowden cable and binding fabric during the transmission process. Brief Description of the Drawings

[0030] Figure 1 is the overall control flowchart for the flexible lower limb assistive exoskeleton during walking;

[0031] Figure 2 is the transmission model diagram of Bowden cables and binding fabrics of any shape;

[0032] Figure 3 is the model diagram of the lower limb and binding fabric in Abaqus;

[0033] Figure 4 is the Fuzzy ADRC control block diagram based on hybrid tension compensation;

[0034] Figure 5 is the fuzzy auto-disturbance rejection control structure diagram;

[0035] Figure 6 is the comparison diagram of the fuzzy auto-disturbance rejection + hybrid tension compensation control and the result without compensation;

[0036] Wherein: 3-1 foot; 3-2 leg binding fabric; 3-3 end binding fabric; 3-4 calf; 3-5 thigh. Detailed Embodiment

[0037] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0038] The present invention deeply analyzes the transmission system of the Bowden cable and binding fabric of the flexible assistive device, from establishing a static Coulomb friction model based on the Bowden cable and a finite element model of the binding fabric, to designing a tension compensation model based on the transmission of the Bowden cable and binding fabric to compensate for the force loss and hysteresis phenomenon during the transmission process. In addition, there is a sign function in the transmission process of the Bowden cable, and considering that there may be other external interference terms and nonlinearities in the Bowden cable and binding fabric system, a fuzzy auto-disturbance rejection control method is adopted to ensure that the output force at the end without a tension sensor can stably follow the input tension, thereby improving the control accuracy of the Bowden cable and binding fabric during the transmission process.

[0039] The control method for Bowden cable and binding fabric transmission provided by the present invention includes the following steps:

[0040] Step 1, construct a flexible lower limb assistive exoskeleton; the flexible lower limb assistive exoskeleton includes a processor, a drive motor, a Bowden cable, a binding fabric, and a sensing and acquisition unit;

[0041] The Bowden cable includes a steel wire rope and a sleeve sleeved outside the steel wire rope; one end of the steel wire rope is connected to the output end of the drive motor; the other end of the steel wire rope is connected to the binding fabric; the binding fabric is fixed on the human leg;

[0042] Step 2, when a person wears the flexible lower limb assistive exoskeleton and walks, the sensing and acquisition unit senses and obtains the human leg movement information in real time, and sends the human leg movement information to the processor;

[0043] Step 3, the processor analyzes the human leg movement information, obtains the current gait cycle, and thus generates an expected pulling force value F0;

[0044] Step 4, the processor uses a hybrid pulling force compensation model to obtain a hybrid pulling force compensation value F3;

[0045] Step 4.1, conduct a micro-element analysis on the Bowden cable drive, establish a static friction model of the Bowden cable, and obtain the friction compensation value F1 of the Bowden cable at the current moment according to the static friction model of the Bowden cable;

[0046] Step 4.1 is specifically as follows:

[0047] The static friction model of the Bowden cable describes the pulling force transmission relationship between the input end and the output end of the Bowden cable. Specifically, when the Bowden cable is stationary, F out = F in (t - ); where, F out is the pulling force at the output end of the Bowden cable at the current moment; F in (t - ) is the pulling force at the input end of the Bowden cable at the previous moment;

[0048] When the Bowden cable moves back and forth, where, is the relative sliding speed of the steel wire rope inside the sleeve, and it is stipulated that the direction consistent with the input pulling force is the positive direction, and the opposite is the negative; μ is the friction coefficient between the steel wire rope and the sleeve, κ(l,t) is the curvature at this place, F out (t - ) is the output pulling force at the previous moment; F in is the pulling force at the input end of the Bowden cable at the current moment; l is the length of the Bowden cable micro-element segment; t is the current time; L is the total length of the Bowden cable.

[0049] Step 4.2, perform numerical simulation analysis on the tensile mechanical properties of the binding fabric by using the finite element method, and establish a tensile compensation model for the binding fabric; according to the tensile compensation model of the binding fabric, obtain the tensile compensation value F2 of the binding fabric at the current moment;

[0050] Specifically, Step 4.2 is as follows:

[0051] Establish a lower limb and binding fabric model; the lower limb and binding fabric model includes a thigh, a calf, a leg binding fabric, and a terminal binding fabric; a hinge connection is provided at the connection between the thigh and the calf; the leg binding fabric is externally bound to the calf; the terminal binding fabric is above the leg binding fabric;

[0052] After dividing the meshes of each component in the lower limb and binding fabric model, use the surface-to-surface binding method to define the interaction between the calf and the leg binding fabric, and the leg binding fabric is connected to the terminal binding fabric by using the binding method;

[0053] Then apply a fully fixed constraint to the thigh, and apply a gravity load to all other components to simulate the gravity received during the walking process;

[0054] Then apply a pressure load in the positive Y direction above the terminal binding fabric, and the magnitude of the load is consistent with the joint moment required by a person during the gait cycle. Therefore, the applied load magnitude is a sine curve with all positive amplitudes, and the periodic amplitude loading is selected to ensure the reciprocating movement of the calf within the period;

[0055] Output the history variables at intervals of 1 / 4 on the upper surfaces of the leg binding fabric and the terminal binding fabric respectively, and then post-process the output results of the history variables of the sectional resultant force. In the post-processed curve, display the change of the output force received by each section;

[0056] Plot the relationship between the load magnitude input from the input end of the binding fabric and the overall average output tensile force received by the terminal binding fabric, obtain the relationship between the input force and the output force of the binding fabric, and further obtain the force transfer model of the binding fabric;

[0057] According to the force transfer model of the binding fabric, obtain the tensile compensation value F2 of the binding fabric at the current moment.

[0058] Step 4.3, perform a summation operation on the friction compensation value F1 of the Bowden wire and the tensile compensation value F2 of the binding fabric to obtain a mixed tensile compensation value F3;

[0059] Step 5, perform a summation calculation on the expected tensile force value F0 and the mixed tensile compensation value F3 to obtain the final tensile input value F4;

[0060] Taking the final tensile input value F4 as the actual input of the system, a fuzzy auto-disturbance rejection controller is adopted to obtain the input tensile force control signal for the driving motor, thereby controlling the driving motor. The driving motor acts on the human leg through a steel wire rope, and the actual tensile force received by the leg is equal to the expected tensile force value F0.

[0061] The principle and design concept of the present invention are introduced below:

[0062] As Figure 1 shown, it is the overall control block diagram of the flexible lower limb assistive exoskeleton during walking. The present invention mainly studies the force transmission part of the exoskeleton, that is, the part within the dotted line box. When the driving motor applies torque, first through Bowden cable transmission, due to the extensibility of the steel cable, the existence of the elongation of the steel cable will cause position errors and motion clearances in the Bowden cable transmission structure. At the same time, due to the friction between the steel cable and the conduit, the Bowden cable transmission structure often encounters problems such as tension attenuation, which hinders the precise control of the Bowden cable transmission. The fabrics of the end-binding fabrics are selected from common polyester webbing and nylon webbing, which are important engineering plastics. However, due to the hysteresis and internal friction characteristics of the fiber fabrics themselves, force loss and hysteresis will occur during the tensile force transmission process of the fabrics, reducing the precise control during the fabric force transmission. Therefore, the present invention provides a good control method for the flexible exoskeleton force transmission system without a force sensor at the end: Bowden cable and binding fabric. The specific solutions are as follows:

[0063] A control method for Bowden cable and binding fabric transmission provided by the present invention consists of two parts: (1) Conduct a micro-element analysis on the Bowden cable transmission, establish a static Coulomb friction model of the Bowden cable, and adopt a finite element method to conduct numerical simulation analysis on the tensile mechanical properties of the binding fabric, thereby establishing a tensile force compensation model; (2) When the mathematical model of the controlled object cannot be accurately described, a fuzzy auto-disturbance rejection controller can be used to estimate and compensate for system disturbances, improving the adaptability and robustness of the control system.

[0064] The force transmission process of the Bowden cable and the binding fabric is mainly realized by the force transmission path composed of a motor, a steel wire rope, a casing, a binding fabric, etc. As Figure 2 shown, one end of the lasso is connected to the fixed motor, and the other end is connected to the binding fabric. The other end of the binding fabric is fixed on the human leg. The dynamic friction transmission model can better describe the friction phenomenon of the lasso system. However, due to the relatively large number of parameters involved in the dynamic friction model and the relatively complex solution of the differential equation, it is not suitable for real-time control. Instead, the "static + Coulomb" friction model is usually adopted to obtain the single-lasso transmission model. Through the micro-element analysis of the lasso, the tensile force transmission relationship between the input end and the output end of the lasso can be obtained: When the Bowden cable is stationary, F out = Fi n (t- ) When the Bowden cable reciprocates, where, is the relative sliding speed of the wire rope inside the casing. It is specified that the positive direction is the same as the direction of the input tensile force, and the negative direction is the opposite. μ is the friction coefficient between the wire rope and the casing, κ(l,t) is the curvature at this point, and F out (t - ) is the output tensile force at the previous moment.

[0065] It is difficult to establish an accurate mathematical model for the force transmission process of the binding fabric. Therefore, the present invention simulates and predicts the mechanical interaction between the binding fabric and the human leg through numerical simulation. The three-dimensional finite element model can establish a lower limb model that conforms to biomechanical theory according to the actual material properties of the human body, improving the accuracy of the simulation. Therefore, the present invention adopts the Abaqus finite element method to study the binding fabric as a macroscopic scale modeling method to analyze the force condition of the end binding fabric. As Figure 3 shown, it is the lower limb and binding fabric model in Abaqus, where 3-1 represents the foot; 3-2 represents the leg binding fabric; 3-3 represents the end binding fabric; 3-4 represents the calf; 3-5 represents the thigh.

[0066] Under the Abaqus software environment, after dividing the meshes of each component of the human lower limb exoskeleton model, the surface-to-surface binding method is adopted to define the interaction between the leg and the binding fabric. The leg binding fabric is also connected to the rectangular binding fabric by the binding method. Then, a complete fixed constraint is applied to the thigh, a gravity load is applied to all other components to simulate the gravity during the walking process, and a hinge connection is set at the connection between the thigh and the calf. Then, a pressure load in the positive Y direction is applied above the rectangular binding fabric. The magnitude of the load should be consistent with the joint moment required by a person during the gait cycle. Therefore, a sine curve with a positive amplitude is used for the applied load magnitude, and the periodic amplitude loading is selected to ensure the reciprocating movement of the calf within the period. By outputting the history variables at intervals of 1 / 4 on the upper surface of the binding fabric and then post-processing the history variable output results of the sectional resultant force, the change of the output force received by each section can be intuitively displayed in the post-processing curve, and the force loss, loss, etc. of the binding fabric during the force transmission process can be analyzed from a macroscopic perspective in all aspects, providing a good characterization for the analysis of the transmission characteristics of the entire Bowden cable and the binding fabric. Then, by using Matlab software to plot the relationship between the load magnitude input from the input end of the binding fabric and the overall average output tensile force received by the end binding fabric, the relationship between the input force and the output force of the binding fabric can be obtained, and then the force transmission model of the binding fabric can be obtained.

[0067] In a human wearable power assist exoskeleton, it is not easy to add a force sensor at the end. Therefore, the present invention designs a hybrid tension compensation controller, which mainly adds the friction compensation value of the Bowden cable and the tension compensation value of the binding fabric to the input tension, so that it does not depend on the end sensing signal, in order to ensure that the output tension can accurately follow the desired trajectory. Therefore, the actual input tension command is the sum of the desired output tension and the compensation tension offset, so as to compensate for the resistance suffered by the Bowden cable-binding fabric. The control principle block diagram is as Figure 4 shown, where only the output information of the driving end of the Bowden cable and binding fabric system is used as the reference output, rather than as feedback acting in the control system. Therefore, the tension compensation control system is mainly composed of a hybrid tension compensation and a fuzzy active disturbance rejection controller.

[0068] PID control is simple to use and convenient to adjust. However, since the PID controller is a linear controller, while the force transmission route of the flexible lower limb power assist exoskeleton in reality is a non-linear system, and there is a sign function in the Bowden cable friction compensation, and there may be other external disturbance terms and non-linearity in the Bowden cable and binding fabric system. Approximating non-linearity with a linear controller will result in a decrease in accuracy. Therefore, an active disturbance rejection controller is adopted on the basis of the hybrid tension compensator. The technology of the Active Disturbances Rejection Controller (ADRC) does not depend on the mathematical model of the controlled object, and can estimate and compensate for system disturbances, improving the adaptability and robustness of the control system. The parameters of the Nonlinear State Error Feedback Control Law (NLSEF) in the ADRC are not easy to adjust. In order to facilitate the actual operation of the ADRC, fuzzy control is introduced on the basis of the ADRC. Utilizing the ability of fuzzy control to perform the best estimation of parameters within a certain range, the gain parameter K in the NLSEF is tuned to achieve the purpose of automatically adjusting the system parameters. The control structure block diagram is as Figure 5 shown. From the simulation results, as Figure 6 shown, the fuzzy active disturbance rejection controller has a fast response and a small overshoot, and has good adaptability to the given tension control. The maximum error and the average value of the absolute error of force tracking are both small. Therefore, the designed method of fuzzy active disturbance rejection controller + feedforward tension compensation can achieve a better control effect, thus realizing the precise tension control of the Bowden cable-binding fabric drive, making the power assist control in the lower limb power assist exoskeleton more precise and effective.

[0069] For example, when a person wears a flexible lower limb assistive exoskeleton and walks normally, IMUs or electromyography sensors attached to the human lower limbs sense and acquire data, which are transmitted to a processor through a communication unit. The processor then determines the current gait cycle based on the acquired signals and inputs a pulling force signal required for the current human gait to the motor. Then, an expected input pulling force signal is generated at the motor end, transmitted through Bowden cables to the transmission of the binding fabric, and finally reaches the human leg. A control method for Bowden cable and binding fabric transmission proposed by the present invention designs a pulling force compensation controller without a sensor at the end. The pulling force compensation controller is jointly determined by factors such as the bending angle and friction coefficient of the Bowden cable and the material properties of the binding fabric. The pulling force compensation of the Bowden cable is determined by analyzing the static Coulomb friction model, mainly determined by the actual bending angle of the Bowden cable, the friction force between the steel wire rope and the sleeve, and the velocity direction of the Bowden cable movement, to obtain a pulling force value that needs to be compensated. The pulling force compensation of the binding fabric is to establish a three-dimensional model of the human lower limb and the binding fabric in the Abaqus finite element analysis software, set and assign material properties to parameters such as the fabric material and size used, define the constraints between the leg and the binding fabric using the surface-to-surface binding method, then mesh each component, and finally apply the required loads and perform post-processing to obtain the relationship between the input and output forces, and thus obtain the pulling force compensation of the binding fabric. The pulling force compensations of the two parts obtained above are integrated, and on this basis, a fuzzy auto-disturbance rejection controller that does not depend on a mathematical model and has adjustable parameters is introduced. Through the hybrid pulling force compensation and fuzzy auto-disturbance rejection control algorithms, the output pulling force received by the human leg is made equal to the expected input pulling force signal of the motor, so as to achieve precise force control without a pulling force sensor at the end, accurately assist human walking, and reduce the metabolic consumption generated by human movement.

[0070] The present invention provides a control method for Bowden cable and binding fabric transmission. Starting from pulling force compensation, it not only considers problems such as dead zones and hysteresis caused by the friction of the Bowden cable, but also takes into account the losses and hysteresis in the force transmission process of the binding fabric, making the force at the end more accurate. Starting from the control method, the method of using a fuzzy auto-disturbance rejection controller with good control adaptability and robustness can significantly improve the force control accuracy and achieve precise force control.

[0071] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also fall within the protection scope of the present invention.

Claims

1. A control method for Bowden wire and binding fabric transmission, characterized in that, It includes the following steps: Step 1, construct a flexible lower limb assistive exoskeleton; the flexible lower limb assistive exoskeleton includes a processor, a drive motor, a Bowden cable, a binding fabric, and a sensing and acquisition unit; The Bowden cable includes a steel wire rope and a sleeve sleeved outside the steel wire rope; one end of the steel wire rope is connected to the output end of the drive motor; the other end of the steel wire rope is connected to the binding fabric; the binding fabric is fixed on the human leg; Step 2, when a person wears the flexible lower limb assistive exoskeleton and walks, the sensing and acquisition unit senses and obtains the human leg movement information in real time, and sends the human leg movement information to the processor; Step 3, the processor analyzes the human leg movement information, obtains the current gait cycle, and thus generates an expected pulling force value F0; Step 4, the processor uses a hybrid pulling force compensation model to obtain a hybrid pulling force compensation value F3; Step 4.1, perform a micro-element analysis on the Bowden cable drive, establish a static friction model of the Bowden cable, and obtain the friction compensation value F1 of the Bowden cable at the current moment according to the static friction model of the Bowden cable; Step 4.1 is specifically: The static friction model of the Bowden cable describes the pulling force transmission relationship between the input end and the output end of the Bowden cable. Specifically, when the Bowden cable is stationary, F out = F in (t - ); where F out is the pulling force at the output end of the Bowden cable at the current moment; F in (t - ) is the pulling force at the input end of the Bowden cable at the previous moment; When the Bowden wire reciprocates, wherein, is the relative sliding speed of the steel wire rope inside the sleeve, and is specified as the positive direction when it is consistent with the direction of the input tensile force, and negative otherwise; μ is the friction coefficient between the steel wire rope and the sleeve, κ(l,t) is the curvature at this point, F out (t - ) is the output tensile force at the previous moment; F in is the tensile force at the input end of the Bowden wire at the current moment; l is the length of the infinitesimal segment of the Bowden wire; t is the current time; L is the total length of the Bowden wire; Step 4.2, perform a numerical simulation analysis of the tensile mechanical properties of the binding fabric by using the finite element method, and establish a tensile force compensation model of the binding fabric; obtain the tensile force compensation value F2 of the binding fabric at the current moment according to the tensile force compensation model of the binding fabric; Step 4.3, perform a summation operation on the friction compensation value F1 of the Bowden cable and the tensile force compensation value F2 of the binding fabric to obtain a hybrid pulling force compensation value F3; Step 5, perform a summation calculation on the expected pulling force value F0 and the hybrid pulling force compensation value F3 to obtain a final pulling force input value F4; Taking the final pulling force input value F4 as the actual input of the system, use a fuzzy auto-disturbance rejection controller to obtain an input pulling force control signal for the drive motor, so as to control the drive motor. The drive motor acts on the human leg through the steel wire rope, and makes the actual pulling force received by the leg equal to the expected pulling force value F0.

2. The control method of Bowden wire and binding fabric drive according to claim 1, wherein Step 4.2 is specifically: Establish a lower limb and binding fabric model; the lower limb and binding fabric model includes a thigh, a calf, a leg binding fabric, and a terminal binding fabric; a hinge connection is provided at the connection between the thigh and the calf; the leg binding fabric is bound outside the calf; the upper part of the leg binding fabric is the terminal binding fabric; After dividing the meshes of each component in the lower limb and binding fabric model, use the surface-to-surface binding method to define the interaction between the calf and the leg binding fabric, and the leg binding fabric is connected to the terminal binding fabric by using the binding method; Then apply a completely fixed constraint to the thigh, and apply a gravity load to all other components to simulate the gravity received during the walking process; Then apply a pressure load in the positive Y direction above the terminal binding fabric, and the magnitude of the load is consistent with the joint torque required by the person in the gait cycle. Therefore, the applied load magnitude is a sine curve with all positive amplitudes, and the periodic amplitude loading is selected to ensure that the calf reciprocates within the period; Output the process variables at 1 / 4 intervals on the upper surfaces of the leg binding fabric and the end binding fabric respectively, and then post-process the output results of the process variables of the sectional resultant force. In the post-processed curve, show the changes in the output force received by each section; Plot the relationship between the load magnitude input from the input end of the binding fabric and the overall average output tension received by the end binding fabric, obtain the relationship between the input force and the output force of the binding fabric, and further obtain the force transfer model of the binding fabric; According to the force transfer model of the binding fabric, obtain the tension compensation value F2 of the binding fabric at the current moment.

Images