A design method of human-machine collaborative controller of lower extremity exoskeleton fusing human fuzzy decision

Abstract

The application discloses a kind of lower limb exoskeleton man-machine collaborative controller design methods of fusing human fuzzy decision, it is related to man-machine interactive control technical field, including: the lower limb exoskeleton dynamics equation of man-machine system containing parameter uncertainty and disturbance is established, and desired trajectory is as constraint and is converted into second-order form;Robust controller is proposed, and stability analysis is carried out to robust controller;A cost functional is proposed, minimize cost functional, obtain optimal fuzzy membership function to realize man-machine cooperation under fuzzy decision;Adjust relevant parameters in system, and analyze the effectiveness of the control method;The collaborative strategy and control method combine human fuzzy decision with mechanical system stability, use human fuzzy decision to regulate the performance of mechanical system in real time to cope with different tasks and working conditions, and the controller has strong robustness and adaptability, and can be used for different control objects.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction control technology, and in particular to a design method for a human-computer collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making. Background Technology

[0002] Human-machine systems, as a core technology integrating the flexibility of human cognition and the precision of machine operation, have been widely applied in key fields such as autonomous driving, industrial collaborative robots, and rehabilitation exoskeletons. Their core advantage lies in making the operator the active subject in the control loop, improving task execution efficiency and adaptability through human-machine collaboration. However, the performance improvement of human-machine systems has long been limited by the technical bottleneck of insufficient human-machine coordination efficiency: on the one hand, machine dynamics itself has inherent nonlinearity, time-varying uncertainty, and system coupling characteristics, making it difficult for traditional control methods to achieve stable tracking of complex task constraints; on the other hand, human decision-making and behavior have significant ambiguity (non-binary characteristics), and their intentional expressions (such as electrophysiological signals, verbal commands, etc.) often lack clear quantitative standards, making them difficult for machines to directly and accurately interpret and integrate.

[0003] To address system uncertainty and constraint tracking problems, constraint tracking control has emerged as a highly promising control paradigm. Its structured characteristics, low energy consumption, and inherent robustness make it suitable for high-dimensional, strongly nonlinear systems. Current research on constraint tracking techniques mainly focuses on two directions: diversifying and expanding constraint types, and handling system uncertainties. However, it fails to fully consider the core characteristic of human-machine systems—the fuzziness of human decision-making—and focuses solely on dynamic optimization at the machine end. This results in insufficient adaptability between control strategies and human intentions, making it difficult to achieve efficient human-machine collaboration in dynamic task scenarios.

[0004] In terms of human intent recognition and integration, existing technologies can capture human symbolic cognitive signals (language, gestures, etc.) and electrophysiological signals (electromyography, electroencephalography, etc.) through sensors, and combine them with intelligent decision-making technologies such as machine learning for intent decoding. However, these methods generally treat human behavior as deterministic input, failing to fully explore and utilize the fuzzy nature of human decision-making. This results in limited accuracy and generalization ability of intent recognition, making it difficult to effectively transform human subjective evaluation and dynamic adjustment needs for tasks into system control parameters. Consequently, the human-machine system lacks adaptability to task changes and cannot maintain optimal performance under complex disturbances and task switching scenarios.

[0005] In existing lower limb exoskeleton control systems, while traditional control methods can ensure system stability to a certain extent, they have significant limitations in balancing human-machine interaction efficiency, uncertainty robustness, and task adaptability. Either they neglect the fuzziness of human behavior, resulting in a poor interactive experience, or they over-rely on fixed control parameters on the machine, leading to a lag in the system's response to task changes, ultimately causing a decline in human-machine collaboration performance and task execution efficiency. Therefore, how to construct a control framework that can accurately integrate human fuzzy decision-making, achieve dynamic adaptation between machine dynamics constraint tracking and human intent expression, and simultaneously ensure system robustness, stability, and optimization performance has become a key technical problem urgently needing to be solved in the field of human-machine systems. Summary of the Invention

[0006] The purpose of this invention is to provide a design method for a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making, thereby solving the problem of insufficient integration of machine dynamics uncertainty handling and human fuzzy decision-making in the prior art.

[0007] To achieve the above objectives, this invention provides a design method for a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making, comprising the following steps: Step 100: Establish the dynamic equations of the human-machine system containing parameter uncertainties and disturbances; Step 200: Use the desired trajectory as a performance constraint and transform it into a second-order form; Step 300: Design a robust controller and perform stability analysis; Step 400: Establish the cost functional and solve for the optimal fuzzy membership function; Step 500: Adjust the system control parameters and verify the effectiveness of the control method.

[0008] Furthermore, step 100 specifically includes the following steps: Step 101, the system dynamic equation of the lower limb exoskeleton is obtained from the Lagrange equation: ; in, It is a generalized coordinate vector. For transpose, , These are the relative joint rotation angles of the hip and knee joints, respectively. It is a generalized velocity vector; It is a generalized acceleration vector; The inertia matrix; Coriolis force / centrifugal force vector; It is the gravity vector; To control the torque vector; Step 102, considering the influence of parameter uncertainties and environmental disturbances on the lower limb exoskeleton system, the system dynamic equations are rewritten as follows: ; in, For uncertain parameter vectors; For unknown environmental interference vectors; The interference input matrix; Step 103: Decompose the matrix and vector containing uncertainty into nominal and uncertain parts, calculated using the following formula: ; ; ; in, , , These are the nominal parts of the inertia matrix, the Coriolis force / centrifugal force vector, and the gravity vector, respectively. , , These are the uncertain parts of the inertia matrix, the Coriolis force / centrifugal force vector, and the gravity vector, respectively.

[0009] Furthermore, the method of using the desired trajectory as a performance constraint and transforming it into a second-order form is as follows: Suppose the system is subject to the following constraints: ; in, For the first In the constraint equations concerning the _th Functional components of a generalized coordinate; For the first The right-hand side terms of each constraint equation; The generalized coordinate dimension of the system; Let be the number of constraint equations, and ; The constraints are represented in matrix form as follows: Taking the first and second derivatives of the matrix constraints respectively, we get: ; ; in, This is the constraint matrix; It is a first-order constraint vector; It is a second-order constraint vector.

[0010] Furthermore, the expression for the designed robust controller is as follows: ; in, For scalar gain parameters, , and These are the upper and lower bounds of the scalar gain parameter's value range, respectively. For nominal control items; To constrain the tracking error processing term; This is an uncertainty compensation item.

[0011] Furthermore, the stability analysis method is as follows: by selecting a Lyapunov function... The derivation yields ,in, Design matrix for positive definite symmetric design. To constrain the tracking error vector, It is a Lyapunov function. Lyapunov function The derivative with respect to time, These are the positive constants associated with the interference input matrix. Let be the true value of the uncertainty bound parameter vector; thus proving the uniform boundedness and uniform eventual boundedness of the controlled system, and the uniform boundedness threshold. and scalar gain parameters The relational expression is: .

[0012] Furthermore, the method for establishing the cost functional is as follows: Step 401, Define Substituting the relationship between the uniform boundedness threshold and the scalar gain parameter, we obtain: ; in, This indicates that the biological signal takes the value of The size of the system's consistent boundedness threshold. , Biosignals generated by humans For biological signal regions, and These are the lower and upper bounds of the biological signal, respectively. The norm of the uncertainty bound parameter vector, For fuzzy membership functions; Step 402, in Define an arc length infinitesimal element in the coordinate system: ; in, It is a differential operator; for Arc length infinitesimal element in a coordinate system; and They are respectively and The differential; Step 403, Quantization function In the interval Total arc length over, construction cost functional: ; in, For cost functionals, for right The derivative of .

[0013] Furthermore, the method used to solve for the optimal fuzzy membership function is as follows: the cost functional is minimized using the variational method to obtain the optimal fuzzy membership function between the scalar gain parameter and the human biological signal; The expression for the optimal fuzzy membership function is as follows: ; in, The optimal fuzzy membership function. , This represents the minimum value of the membership function; , This is the midpoint of the biological signal interval.

[0014] Furthermore, the optimal fuzzy membership function obtained by solving satisfies the following properties: continuity, symmetry about the midpoint of the biosignal interval, and value at the boundary. and the value at the midpoint .

[0015] Furthermore, the system control parameters that need to be adjusted are: and .

[0016] Furthermore, the method to verify the effectiveness of the control method is to perform simulation verification using the standard gait trajectory as the constraint of the desired trajectory under the presence of uncertainties and unknown disturbances.

[0017] Therefore, the present invention employs the above-mentioned design method for a lower limb exoskeleton human-machine collaborative controller that integrates human fuzzy decision-making, and has the following beneficial effects: (1) The robust controller designed includes nominal control terms, constraint tracking error processing terms and uncertainty compensation terms, which can effectively cope with system parameter uncertainty and unknown environmental disturbances, and ensure the stable operation of the system under complex working conditions. Furthermore, the uniform boundedness and uniform final boundedness of the controlled system are rigorously proved by Lyapunov stability theory.

[0018] (2) By constructing a cost functional and using variational method to perform a minimization operation, an optimal fuzzy membership function with a closed solution form is obtained. This function can transform the fuzzy decision made by the wearer after perceiving changes in the environment into the optimal gain parameter in the controller, thus avoiding the performance of the lower limb exoskeleton being too affected by the fuzziness of human decision-making, and completing the task under the current environment and conditions with optimal performance.

[0019] (3) Both the optimal fuzzy membership function and the cost functional are closed-form solutions. Their explicit mathematical forms can ensure sufficient accuracy required for practical engineering applications. Compared with numerical methods, they can support more rigorous theoretical analysis. The membership function has global uniqueness, which indicates the existence of a deterministic artificial behavior pattern that is not affected by factors such as models and uncertainties.

[0020] (4) The human fuzzy decision-making is organically combined with the robust control of the mechanical system. On the one hand, it relies on online state feedback and robust control to deal with the uncertainty of machine dynamics and environmental disturbances in real time. On the other hand, it collects human biological signals through sensors and processes them through the optimal membership function to transform human fuzzy decision-making into dynamic gain parameters and feeds them back to the controller, forming an efficient human-machine collaborative strategy. The controller has strong robustness and adaptability.

[0021] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0022] Figure 1 This is a flowchart of a design method for a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making, according to the present invention. Figure 2 This is a schematic diagram of the lower limb exoskeleton system of the present invention; Figure 3 This is a schematic diagram of the overall structure of the robust controller of the present invention; Figure 4 This is the optimal fuzzy membership function graph of the present invention; Figure 5 This is a simulation diagram of the robust controller of the present invention, wherein (a) is a diagram of the angle change of joint 1 and (b) is a diagram of the angle change of joint 2. Detailed Implementation

[0023] The following detailed description of embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely illustrates selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0024] Please see Figure 1A design method for a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making includes the following steps: Step 100: Establish the dynamic equations of the human-machine system containing parameter uncertainties and disturbances; Please see Figure 2 The lower limb exoskeleton system is a two-link model with two joints (hip and knee joints), using the joint rotation angles as controlled generalized variables to simulate the gait trajectory of a human walking limb. The system dynamics equations of the lower limb exoskeleton, derived from the Lagrange equations, are as follows: ; In the formula: It is a generalized coordinate vector. For transpose, , These are the relative joint rotation angles of the hip and knee joints, respectively. It is a generalized velocity vector; It is a generalized acceleration vector; The inertia matrix, abbreviated as ; The Coriolis force / centrifugal force vector, abbreviated as ; The gravitational vector, abbreviated as ; To control the torque vector, it is abbreviated as .

[0025] Specifically, the element expressions for each matrix and vector are as follows: , , , ; The elements of the inertia matrix are: ; ; ; ; The elements of the Coriolis force / centrifugal force vector are: ; ; The elements of the gravity vector are: ; ; In the above formulas, This refers to the length of link 1 (thigh bar); This is the length of link 2 (lower leg); This is the distance from the center of mass of link 1 to the origin of the hip joint; This is the distance from the center of mass of link 2 to the knee joint; Let the mass of link 1 be denoted as 'Mass'. Let the mass of link 2 be denoted as '2'. For the quality of the hip joint module; To determine the mass of the knee joint module, the module at joint 2 is treated as a point mass. Let be the moment of inertia of link 1; Let be the moment of inertia of link 2; The moment of inertia of the hip joint module; It is the acceleration due to gravity; , , , , , .

[0026] It should be noted that in actual human-machine systems, due to manufacturing errors, load variations, environmental disturbances, and other factors, the inertia matrix, Coriolis force / centrifugal force vector, and gravity vector of the lower limb exoskeleton system all contain uncertainties. Considering the influence of parameter uncertainties and environmental disturbances on the lower limb exoskeleton system, equation (1) is rewritten as: ; in, For uncertain parameter vectors; For unknown environmental interference vectors; This is the interference input matrix.

[0027] Decompose matrices and vectors containing uncertainty into nominal and uncertain parts: ; ; ; in, , , These are the nominal parts of the inertia matrix, Coriolis force / centrifugal force vector, and gravity vector, respectively, representing the known parts corresponding to the nominal values ​​of the system parameters; , , These represent the uncertainties in the inertia matrix, the Coriolis force / centrifugal force vector, and the gravity vector, respectively, indicating deviations caused by parameter uncertainties.

[0028] Step 200: Use the desired trajectory as a performance constraint and transform it into a second-order form; To ensure the lower limb exoskeleton system moves along a desired trajectory, the desired trajectory is used as a performance constraint for the system. Assume the system is subject to the following constraints: ; in, For the first In the constraint equations concerning the _th Functional components of a generalized coordinate; For the first The right-hand side terms of each constraint equation; The generalized coordinate dimension of the system; Let be the number of constraint equations, and ; and All are first-order differentiable functions.

[0029] Equation (2) can be expressed in matrix form as follows: ; in, The constraint function vector; To constrain the vector on the right.

[0030] Taking the first and second derivatives of equation (3) respectively, we can obtain the first and second order forms of the constraint: ; ; in, The constraint matrix is ​​abbreviated as From equation (3) on the generalized coordinates Taking the partial derivative yields the result; A first-order constraint vector, abbreviated as ; It is a second-order constraint vector, abbreviated as .

[0031] In embodiments of the present invention, for a lower limb exoskeleton system, the following standard gait trajectory constraints for simulating walking are considered: ; ; in, , Hip joint rotation angle and knee joint angle The expected trajectory is time. The Fourier series form of the coefficients corresponds to the amplitudes of different frequency components.

[0032] Therefore, the constraint matrix and constraint vector of the lower limb exoskeleton system can be obtained as follows: ; in, , Desired trajectories , The first derivative with respect to time; , These are the second derivatives of the desired trajectory with respect to time.

[0033] Step 300: Design a robust controller and perform stability analysis; Please see Figure 3 Based on the system dynamics equations with uncertainties obtained in step S1 and the second-order form of the constraints obtained in step S2, the following robust controller is designed: ; in, This is a scalar gain parameter, and its value depends on human fuzzy decision-making. For humans, This is the optimal result of fuzzy decision-making; for machines, These are the optimal control parameters for manipulating the robot to adapt to the current task, and the specific method for determining them will be introduced in step S4. For nominal control items; To constrain the tracking error processing term; This is an uncertainty compensation term. The specific expressions for each term are as follows: ; in, for The square root of the matrix, due to Since it is a positive definite symmetric matrix, its square root exists and is unique. for The inverse matrix; Represents the Moore-Penrose generalized inverse; It is a second-order constraint vector; This is the nominal part of the Coriolis force / centrifugal force vector; This is the nominal part of the gravity vector. Its function is to achieve accurate tracking of constraints under the nominal model.

[0034] ; in, The design matrix is ​​positive definite and symmetric. , The constraint tracking error vector represents the deviation between the actual motion state of the system and the desired constraint. Its function is to drive the system state to converge toward the constraint surface when the initial conditions do not meet the performance constraints.

[0035] ; in, For the estimated parameter vector of the uncertainty bound; The uncertainty influence function vector is constructed from the known information of the system; It is a unified upper bound scalar for the impact of system uncertainty on constraint tracking error. Its function is to compensate for the impact of system uncertainties on constraint tracking performance.

[0036] Next, to verify that the controlled system can achieve uniform boundedness and uniform eventual boundedness, a stability analysis is performed on the robust controller. The Lyapunov function is selected as follows: ; in, It is a Lyapunov function. To constrain the tracking error vector, The design matrix is ​​positive definite and symmetric. Because... It is a positive definite matrix, therefore ,and If and only if .

[0037] By taking the time derivative of the Lyapunov function and substituting it into the robust controller expression, we can derive and calculate the following: ; in, Lyapunov function The derivative with respect to time; To the interference input matrix The relevant positive constants reflect the degree of impact of environmental disturbances on the system; The true value of the uncertainty bound parameter vector; This represents the Euclidean norm.

[0038] From equation (4), it can be seen that when When large enough, This proves that the controlled system possesses uniform boundedness. Specifically, the uniformly bounded function... for: ; in, Let be the norm of the initial state of the system; The uniformity-boundedness threshold is defined as: ; and Let the constants related to the Lyapunov function be defined as follows: ; ; in, and These are positive definite symmetric design matrices. The minimum and maximum eigenvalues; This is the positive design constant associated with uncertainty estimation.

[0039] Furthermore, the controlled system also possesses uniform eventual boundedness. The final bound of uniform eventual boundedness is: ; The upper bound of the time required for convergence from the initial state to the final bound is: ; in, Let be the radius of the given bounded region of the target.

[0040] It should be noted that, as shown in equation (5), the uniformity boundedness threshold... With scalar gain parameter They are inversely proportional. That is... The larger the threshold The smaller the value, the smaller the region where the system's constraint tracking error can eventually converge, and the higher the control accuracy. This relationship constitutes the basis for setting the scalar gain parameter in step S4. The theoretical basis for linking human fuzzy decision-making.

[0041] Step 400: Establish the cost functional and solve for the optimal fuzzy membership function; To achieve scalar gain parameters for human fuzzy decision-making and robust controllers The optimal mapping between them is proposed. The selection method is as follows: ; in, , and These are the upper and lower bounds of the scalar gain parameter's value range, respectively. , Biological signals are those generated by humans. These signals can be electrical signals (such as electromyography and electroencephalography), mechanical signals (such as force / torque signals), chemical signals, or physiological signals produced by the organism. For biological signal regions, ; The fuzzy membership function is the medium that transforms fuzzy human biological signals into optimal control parameters in a robust controller. When... Once selected, it will remain unchanged throughout the entire task execution until it is re-determined when switching to the next task.

[0042] To determine the fuzzy membership function The optimal form, consider a membership function And it is required to have the following properties: (1) Continuity: In the domain The above is a continuous function; (2) Symmetry: Relative to the midpoint of the biosignal interval Symmetry, that is, for any ,have ; (3) Boundary condition 1: When or hour, ,in This represents the minimum value of the membership function; (4) Boundary condition 2: When hour, ,at this time Take the maximum value.

[0043] Based on the uniform boundedness threshold in equation (5) and The relationship will Substitution The expression is defined as follows: ; in, This indicates that the biological signal takes the value of The size of the system's consistent boundedness threshold. The norm of the uncertainty bound parameter vector, This is the upper bound of the scalar gain parameter. The fuzzy membership function to be determined.

[0044] exist In the coordinate system, define the arc length infinitesimal element: ; in, It is a differential operator; for Arc length infinitesimal element in a coordinate system; and They are respectively and The differential. From equation (6), we can obtain: ; in, , for right The derivative of .

[0045] Define the following cost functional to quantize the function. In the interval Total arc length on: ; in, The cost functional, in its physical sense, reflects the effect of a system's uniformly bounded and uniformly eventually bounded thresholds on biological signals. Sensitivity to change. Smaller. Indicates system stability The system has low sensitivity to changes, meaning that human fuzzy decisions have a relatively small impact on the system's stability. Therefore, to minimize the impact of human decision-making fuzziness on system performance, it is necessary to minimize the cost functional.

[0046] By minimizing equation (7) using the variational method, and under the constraints of the aforementioned four properties, the following optimal fuzzy membership function can be obtained. The expression is as follows: ; in, , This represents the minimum value of the membership function; , This is the midpoint of the biological signal interval; and These are the lower and upper bounds of the biological signal, respectively.

[0047] It should be noted that the optimal fuzzy membership function has the following properties: (a) when hour, ,correspond ; (b) When hour, ,correspond ; (c) When hour, ,correspond .

[0048] Membership function with respect to midpoint It is symmetric, continuous, and satisfies all the pre-defined property conditions.

[0049] The optimal membership function is a crucial design element in human-machine collaborative strategies, transforming the fuzzy decisions made by the wearer after perceiving changes in the environment into optimal gain parameters in a robust controller. This avoids the performance of the lower limb exoskeleton being excessively affected by the ambiguity of human decision-making, while achieving optimal performance to complete tasks under the current environment and conditions.

[0050] Both the cost functional and the membership function are constructed in closed-form solution. These analytical expressions have two core advantages: first, the explicit mathematical form ensures sufficient accuracy for practical engineering applications; second, compared to numerical methods, they support more rigorous theoretical analysis. As an intermediary between optimal and fuzzy decision-making, the membership function possesses global uniqueness, indicating the existence of a deterministic pattern of human behavior unaffected by factors such as models and uncertainties.

[0051] In an embodiment of the present invention, Figure 4 The figure shows the optimal fuzzy membership function obtained after minimizing the cost functional. It serves as a bridge for transforming human fuzzy decision-making into controller gain parameters, enabling the controller to adjust system performance in real time based on changes in biological signals to adapt to the environment and specific tasks.

[0052] Step 500: Adjust the system control parameters and verify the effectiveness of the control method.

[0053] The control parameters that need to be adjusted in the robust controller of the human-machine system that incorporates human fuzzy decision-making are: and Its selection aims to define the scalar gain parameter. effective range This provides clear boundary constraints for fuzzy decision-making in humans. Specifically: upper bound of the interval The optimal control performance that the system can achieve has been determined, that is, when human biological signals correspond to... hour, At this point, the system's constraint tracking error finally reaches its limit. Minimum, highest control precision.

[0054] Lower bound of the interval This determines the minimum control performance guarantee of the system, that is, when human biological signals are at the boundary value. or hour, The system can still maintain basic stability and constraint tracking capabilities.

[0055] The size of the interval is jointly determined by the human decision-making scope and the optimal fuzzy membership function. Combining the optimal membership function designed based on the variational method ensures... The system should always remain within the specified range, preserving the flexibility of human decision-making while preventing performance from exceeding acceptable limits due to decision ambiguity. The specific values ​​for the upper and lower bounds need to be selected by the designers based on the actual situation.

[0056] After confirming the control parameters, perform control performance simulation and evaluate the simulation results to check whether they meet the preset control requirements (such as tracking accuracy, robustness indicators, response speed, etc.). If the simulation results meet all control requirements, the control design process ends; otherwise, if the requirements are not met, the controller parameters need to be readjusted and simulation verification needs to be performed again. This process is iterated until the control performance meets the standards.

[0057] Please see Figure 5 Figure (a) shows the angle change of joint 1, and figure (b) shows the angle change of joint 2. After adjusting the necessary parameters in the robust controller, simulation verification was performed under the condition of uncertainty and unknown disturbance, with the standard gait trajectory as the constraint of the desired trajectory (corresponding to the dashed lines in figures (a) and (b)). As shown in the simulation results, when the human biological signal values ​​are 13.19, 11.70, 7.80, and 3.35, the designed robust controller can effectively overcome the influence of unknown interference and system uncertainty, driving the actual joint angle changes to accurately follow the desired trajectory with the error within the allowable range. Moreover, it maintains relatively consistent tracking performance under different signal values, fully demonstrating the excellent robustness and dynamic response capability of the control strategy, thus strongly verifying its feasibility and effectiveness in the actual control of exoskeleton robots.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A design method for a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making, characterized in that, Includes the following steps: Step 100: Establish the dynamic equations of the human-machine system containing parameter uncertainties and disturbances; Step 200: Use the desired trajectory as a performance constraint and transform it into a second-order form; Step 300: Design a robust controller and perform stability analysis; Step 400: Establish the cost functional and solve for the optimal fuzzy membership function; Step 500: Adjust the system control parameters and verify the effectiveness of the control method; The method for establishing the cost functional is as follows: Step 401, Define Substituting the relationship between the uniform boundedness threshold and the scalar gain parameter, we obtain: ； in, This indicates that the biological signal takes the value of The size of the system's consistent boundedness threshold. , Biosignals generated by humans For biological signal regions, and These are the lower and upper bounds of the biological signal, respectively. The norm of the uncertainty bound parameter vector, For fuzzy membership functions; For scalar gain parameters, , and These are the upper and lower bounds of the scalar gain parameter's value range, respectively. Step 402, in Define an arc length infinitesimal element in the coordinate system: ； in, It is a differential operator; for Arc length infinitesimal element in a coordinate system; and They are respectively and The differential; Step 403, Quantization function In the interval Total arc length over, construction cost functional: ； in, For cost functionals, for right The derivative; The method used to solve for the optimal fuzzy membership function is as follows: the cost functional is minimized using the variational method to obtain the optimal fuzzy membership function between the scalar gain parameter and the human biological signal. The expression for the optimal fuzzy membership function is as follows: ； in, The optimal fuzzy membership function. , This represents the minimum value of the membership function; , This is the midpoint of the biological signal interval.

2. The design method of a human-machine collaborative controller for a lower limb exoskeleton that integrates human fuzzy decision-making as described in claim 1, characterized in that, Step 100 specifically includes the following steps: Step 101, the system dynamic equation of the lower limb exoskeleton is obtained from the Lagrange equation: ； in, It is a generalized coordinate vector. For transpose, , These are the relative joint rotation angles of the hip and knee joints, respectively. It is a generalized velocity vector; It is a generalized acceleration vector; The inertia matrix; Coriolis force / centrifugal force vector; It is the gravity vector; To control the torque vector; Step 102, considering the influence of parameter uncertainties and environmental disturbances on the lower limb exoskeleton system, the system dynamic equations are rewritten as follows: ； in, For uncertain parameter vectors; For unknown environmental interference vectors; The interference input matrix; Step 103: Decompose the matrix and vector containing uncertainty into nominal and uncertain parts, calculated using the following formula: ； ； ； in, , , These are the nominal parts of the inertia matrix, the Coriolis force / centrifugal force vector, and the gravity vector, respectively. , , These are the uncertain parts of the inertia matrix, the Coriolis force / centrifugal force vector, and the gravity vector, respectively.

3. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 2, characterized in that, The method of using the desired trajectory as a performance constraint and transforming it into a second-order form is as follows: Suppose the system is subject to the following constraints: ； in, For the first In the constraint equations concerning the _th Functional components of a generalized coordinate; For the first The right-hand side terms of each constraint equation; The generalized coordinate dimension of the system; Let be the number of constraint equations, and ; The constraints are represented in matrix form as follows: Taking the first and second derivatives of the matrix constraints respectively, we get: ； ； in, This is the constraint matrix; It is a first-order constraint vector; It is a second-order constraint vector.

4. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 3, characterized in that, The expression for the designed robust controller is as follows: ； in, For nominal control items; To constrain the tracking error processing term; This is an uncertainty compensation item.

5. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 4, characterized in that, The stability analysis method is as follows: by selecting Lyapunov functions The derivation yields ,in, Design matrix for positive definite symmetric design. To constrain the tracking error vector, It is a Lyapunov function. Lyapunov function The derivative with respect to time, These are the positive constants associated with the interference input matrix. Let be the true value of the uncertainty bound parameter vector; thus proving the uniform boundedness and uniform eventual boundedness of the controlled system, and the uniform boundedness threshold. and scalar gain parameters The relational expression is: .

6. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 5, characterized in that, The optimal fuzzy membership function obtained by solving satisfies the following properties: continuity, symmetry about the midpoint of the biosignal interval, and value at the boundary. and the value at the midpoint .

7. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 6, characterized in that, The system control parameters that need to be adjusted are: and .

8. The design method of a lower limb exoskeleton human-machine collaborative controller integrating human fuzzy decision-making as described in claim 7, characterized in that, The method to verify the effectiveness of the control method is to perform simulation verification under the condition of uncertainty and unknown disturbance, using the standard gait trajectory as the constraint of the desired trajectory.

Images