Lower limb exoskeleton variable stiffness compliance control method based on finite time interference prediction

Abstract

The invention relates to a lower limb exoskeleton robot control technology, and discloses a variable stiffness compliance control method based on finite time interference prediction. The method comprises the following steps: establishing a lower limb hip and knee joint sagittal plane dynamic model containing human-computer interaction force and external interference; designing an impedance model based on the relationship between the man-machine interaction force and the static balance force, and outputting a reference angle track; constructing an outer ring virtual control law; constructing a finite time interference observer, and identifying unmeasurable interference on line; and designing the control method according to a feedback linearization principle by utilizing the result. According to the method, the response flexibility of the hip and knee joints of the lower limb exoskeleton can be improved, the man-machine coordination problem caused by sudden increase of man-machine interaction force due to sudden movement change is avoided, the joint injury risk of a wearer is eliminated, the control robustness is enhanced, and real-time trajectory tracking is achieved.

Description

Technical Field

[0001] The present application belongs to the field of lower limb exoskeleton robot control technology, and specifically relates to a variable stiffness compliant control method for a lower limb exoskeleton based on finite time interference prediction. Background Art

[0002] In recent years, with the rapid advancement of technologies in ergonomics, mechanical design, and electrical systems, lower-limb exoskeletons, as innovative wearable devices, have become a growing focus of attention in both research and industry. The primary design objective of these exoskeletons is to form a close, coordinated motion relationship with the human body, providing the wearer with the necessary auxiliary power support. This significantly improves the wearer's endurance, enabling them to maintain optimal performance during prolonged or high-intensity tasks.

[0003] However, achieving efficient, precise, and compliant control of lower-limb exoskeleton robots has always been a key challenge in the development of this field. Currently, traditional lower-limb exoskeleton control methods, such as position-based and force-based control strategies, can meet basic application requirements to a certain extent, but there are still two major problems that need to be solved:

[0004] On the one hand, the interaction between the wearer and the exoskeleton robot is a factor that cannot be ignored during human-robot collaborative motion. Traditional control methods often lack effective interference identification and compensation mechanisms, and can only achieve partial optimization by adjusting the parameters of the underlying controller. This cannot fundamentally solve the problems of reduced control accuracy and stability caused by interaction interference.

[0005] On the other hand, current mainstream lower-limb exoskeleton robot motion trajectory design often uses a pre-programmed approach, where the robot's motion path and speed are pre-set. However, due to the variability of the wearer's environment and individual gait differences, these pre-programmed motion trajectories often fail to fully meet the wearer's actual needs in practical applications, resulting in insufficient compliance and robustness during the exoskeleton's tracking and control process. Summary of the Invention

[0006] In view of the limitations of existing technologies in solving the problem of high-compliance and strong robust tracking control of exoskeleton robots under multi-source interference, this application proposes a variable-stiffness compliance control method for lower limb exoskeleton based on finite-time interference prediction. This method aims to improve the compliance response capability of the lower limb exoskeleton and the robustness of the overall control system by online identification and compensation of model parameter uncertainties and external interference.

[0007] In order to achieve the above technical objectives, this application mainly adopts the following technical solutions:

[0008] In an aspect of the application, a variable stiffness compliant control method for lower extremity exoskeleton based on finite-time disturbance prediction is provided, comprising the following steps:

[0009] A lower extremity hip-knee joint sagittal plane motion dynamics model is established, which represents human-robot interaction force and external disturbance.

[0010] An impedance model is designed based on the dynamic relationship between human-robot interaction force and static balance force, and a reference angle trajectory under flexible target guidance is outputted by the impedance model based on the desired angle trajectory;

[0011] A virtual control law of an outer loop angle circuit is constructed based on the reference angle trajectory and the current motion angle state, which is used to generate a control signal to guide the motion of the lower extremity exoskeleton to tend to the reference angle trajectory;

[0012] A finite-time disturbance observer is constructed based on the lower extremity exoskeleton hip-knee joint sagittal plane motion dynamics model, and the unmeasurable disturbance information in the motion process is identified online by the finite-time disturbance observer;

[0013] Based on the feedback linearization principle, a variable stiffness compliant control method for lower extremity exoskeleton is designed by using the design results of the virtual control law and the finite-time disturbance observer, which adjusts the stiffness of the lower extremity exoskeleton in real time according to the unmeasurable disturbance information identified online and the control signal generated by the virtual control law to realize compliant control.

[0014] In an embodiment, a lower extremity hip-knee joint sagittal plane motion dynamics model is established based on the Lagrange modeling method, which represents human-robot interaction force and external disturbance:

[0015]

[0016] Wherein, M0(q) represents a positive definite symmetric inertia matrix, is a centrifugal force and Coriolis force matrix, G0(q) represents a gravity matrix, q, respectively represent the angle, angular velocity and angular acceleration of the hip-knee joint of the lower extremity exoskeleton; ΔM(q), ΔG(q) is the uncertainty of model parameter identification, τ is the control torque input, τ h is the human-robot interaction force of the assistant, d is the external unmeasurable disturbance in the motion process;

[0017] The constructed dynamics model is simplified and combined:

[0018]

[0019] Wherein, represents the lumped disturbance of the lower extremity exoskeleton control system including model parameter uncertainty and external disturbance.

[0020] Define x1 = q and Design the sagittal kinematic / dynamic model of the lower limb hip and knee joints in state space form:

[0021]

[0022] In one embodiment, based on the dynamic relationship between human-machine interaction force and static balance force, an impedance model based on stiffness and damping is constructed to output a reference angle trajectory under flexible target guidance:

[0023]

[0024] Among them, τ b represents the static equilibrium force, B and K represent the positive definite damping and stiffness matrices respectively, x d Defined as the desired angle trajectory, x c Represents the reference angle trajectory of the impedance model output.

[0025] In one embodiment, a virtual control law of the outer angle loop is designed based on the reference angle trajectory using the feedback principle:

[0026]

[0027] Where k1 is the control gain of the outer angle loop.

[0028] In one embodiment, the finite-time disturbance observer is constructed based on the finite-time convergence stability theorem according to the sagittal kinematics / dynamics model of the lower limb hip and knee joints:

[0029]

[0030] in, Represent the real-time angle, angular velocity and interference estimation results, λ1=diag{λ 1,1 ,λ 1,2},λ2=diag{λ 2,1 ,λ 2,2}, is the observer design parameter, 0<y<1, To design a function, its specific form is designed as follows:

[0031]

[0032] In one embodiment, the design results of the virtual control law and the disturbance observer are used to construct a lower limb hip and knee joint sagittal plane motion dynamics input torque controller based on the feedback linearization principle:

[0033]

[0034] wherein k2 is an input torque controller gain.

[0035] In another aspect of the present application, a lower extremity exoskeleton control device is provided, characterized by comprising a processor and a memory, the memory storing a computer program, and the processor implementing the lower extremity exoskeleton variable stiffness compliance control method as described above when executing the computer program.

[0036] In another aspect of the present application, a computer readable storage medium is provided, characterized by storing a computer program, and the computer program implementing the lower extremity exoskeleton variable stiffness compliance control method as described above when executed by a processor.

[0037] The present application has the following beneficial effects:

[0038] By constructing a finite time disturbance observer, online identification and estimation of model parameter uncertainty and external disturbance are achieved, and finite time convergence of disturbance estimation error is achieved, thereby providing state input for real-time compensation;

[0039] By introducing an impedance model, reference angle trajectory generation under the compliance target guidance is achieved under the dynamic relationship between human-machine interaction force and static balance force, and the compliance response capability of the lower extremity hip and knee joints is improved;

[0040] By constructing a virtual control law of the outer loop angle circuit and a dynamic input torque controller, and compensating the lumped disturbance in the input torque controller, the robustness and anti-interference ability of the overall control system are improved. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 Fig. 1 is a schematic diagram of a lower extremity exoskeleton model in an embodiment of the present application;

[0042] Figure 2 Fig. 2 is a block diagram of the control method proposed in an embodiment of the present application;

[0043] Figure 3 Fig. 3 is a trajectory tracking schematic diagram of the control method proposed in an embodiment of the present application;

[0044] Figure 4 Fig. 4 is a schematic diagram of the comparison between the lumped disturbance estimation value and the true value of the control method proposed in an embodiment of the present application;

[0045] Figure 5 Fig. 5 is a schematic diagram of the angle trajectory adjustment effect of different stiffness of the control method proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0047] In one embodiment of the present application, a method for variable stiffness compliance control of a lower limb exoskeleton based on finite time interference prediction is provided, comprising the following steps:

[0048] S1: Establish a dynamic model for the sagittal plane motion of the lower limb hip and knee joints that includes human-machine interaction forces and external disturbances.

[0049] When establishing a sagittal plane motion dynamics model of the lower limb hip and knee joints that can characterize human-machine interaction forces and external interference, factors such as the mass, inertia, joint friction, human-machine interaction forces, and external interference forces of the lower limb exoskeleton are comprehensively considered. The dynamic equations can be established using the Lagrange equation or the Newton-Euler method to accurately describe the motion laws of the lower limb exoskeleton hip and knee joints in the sagittal plane.

[0050] In certain embodiments, based on the Lagrangian modeling method, a sagittal plane motion dynamics model of the lower limb hip and knee joints is established to characterize human-machine interaction forces and external interference, taking into account model uncertainty, mismatch, etc.:

[0051]

[0052] Where M0(q) represents the positive definite symmetric inertia matrix; is the centrifugal force and Coriolis force matrix; G0(q) represents the gravity matrix; q, Respectively represent the angle, angular velocity and angular acceleration of the lower limb exoskeleton hip and knee joints; ΔM(q), ΔG(q) is the uncertainty of model parameter identification, which is the deviation between the actual inertia matrix and the inertia matrix assumed in the model, the deviation between the actual centrifugal force and Coriolis force matrices and the matrices assumed in the model, and the deviation between the actual gravity matrix and the gravity matrix assumed in the model; τ is the control torque input; τ h is the auxiliary human-machine interaction force; d is the external unmeasurable interference during the motion process;

[0053] Simplify and merge the constructed dynamic models:

[0054]

[0055] in,

[0056] Furthermore, define x1 = q and Design the sagittal kinematic / dynamic model of the lower limb hip and knee joints in state space form:

[0057]

[0058] in, It represents the lumped disturbance of the lower limb exoskeleton control system including model parameter uncertainty and external disturbance, is the rate of change of x1; is the rate of change of x2.

[0059] S2: Based on the dynamic relationship between the human-machine interaction force and the static balance force, an impedance model is designed on the basis of the desired angle trajectory, and the reference angle trajectory under the flexible target guidance is output through the impedance model.

[0060] When designing the impedance model based on the dynamic relationship between the human-computer interaction force and the static balance force, the mapping relationship between the human-computer interaction force and the static balance force is determined through experiments or simulation analysis, and this mapping relationship is introduced into the impedance model, so that the impedance model can dynamically adjust the reference angle trajectory according to the real-time human-computer interaction force to achieve a flexible goal-oriented control effect.

[0061] In some embodiments, an impedance model based on stiffness and damping is constructed to achieve the transformation from the desired angle trajectory to the reference angle trajectory:

[0062]

[0063] Among them, τ b represents the static equilibrium force; B and K represent the positive definite damping and stiffness matrices; x d Defined as the desired angle trajectory; x c represents the reference angle trajectory of the impedance model output; is the desired angle trajectory x d The derivative of , represents the desired angular velocity trajectory; is the reference angle trajectory x c The derivative of , represents the reference angular velocity trajectory.

[0064] S3: Constructing a virtual control law of the outer angle loop according to the reference angle trajectory and the current motion angle state, wherein the virtual control law is used to generate a control signal to guide the movement of the lower limb exoskeleton toward the reference angle trajectory.

[0065] When constructing the virtual control law of the outer angle loop, methods such as PID control, sliding mode control, or adaptive control are used to generate a control signal based on the deviation between the reference angle trajectory and the current motion angle state. This control signal serves as the input of the inner loop control to guide the movement of the lower limb exoskeleton.

[0066] Reference angle trajectories can be pre-planned based on different application scenarios and control objectives. For example, in rehabilitation training, a series of angular trajectories of increasing difficulty can be designed based on the patient's recovery stage and motor ability. In assisted walking, angular trajectories that conform to human kinematic characteristics can be generated based on the gait patterns of normal human walking. Reference angle trajectories can be stored in the control system's memory through offline planning or generated online based on real-time human-computer interaction or environmental information.

[0067] To ensure the stability, response speed, and accuracy of the control system, the parameters of the virtual control law can be adjusted and optimized. Trial and error methods, empirical formulas, and intelligent optimization algorithms (such as genetic algorithms and particle swarm optimization) can be used to determine the optimal parameter values. Taking PID control as an example, through trial and error, the proportional, integral, and differential parameters are gradually adjusted, and the control system response curve is observed until the design requirements are met.

[0068] In some embodiments, the feedback principle is used to design a virtual control law for the outer angle loop:

[0069]

[0070] Among them, α is the virtual control quantity, the output of the outer loop controller, which serves as the reference input of the inner loop; k1 is the control gain of the outer loop angle loop.

[0071] S4: Constructing a finite-time disturbance observer based on the sagittal motion dynamics model of the lower limb exoskeleton hip and knee joints, and using the finite-time disturbance observer to online identify unmeasurable disturbance information during the motion process.

[0072] Based on the sagittal plane dynamics model of the hip and knee joints in a lower-limb exoskeleton, the structure and parameters of a finite-time disturbance observer are designed to accurately estimate unmeasurable disturbance information within a finite time. During the design process, finite-time convergence control theories and methods, such as finite-time stability theory and sliding mode control techniques, are employed to ensure that the disturbance observer's estimation error converges to zero or a small neighborhood near zero within a finite time.

[0073] The actual motion data of the lower limb exoskeleton (such as joint angles, angular velocities, torques, etc.) are input into the constructed finite-time disturbance observer. Based on the input actual motion data and the dynamic model of the lower limb exoskeleton, the finite-time disturbance observer calculates and outputs the estimated value of the unmeasurable disturbance information in real time, including external disturbance forces, disturbances caused by model uncertainty, etc.

[0074] In certain embodiments, according to the lower limb hip-knee joint sagittal plane kinematics / dynamics model in the state space form, a finite time disturbance observer is constructed based on the finite time convergence stability theorem to realize online identification of the unmeasurable disturbance information during the motion:

[0075]

[0076] wherein, respectively represent the real-time angle, angular velocity, and the estimated result of the lumped disturbance; λ1 = diag{λ 1,1 , λ 1,2}, λ2 = diag{λ 2,1 , λ 2,2} are the observer design parameters; are respectively the estimated value of the lower limb exoskeleton joint angular velocity, the estimated value of the angular acceleration, and the estimated value of the lumped disturbance rate of change; 0 < y < 1; is a design function, and the specific form is designed as:

[0077]

[0078] S5: using the design results of the virtual control law and the finite time disturbance observer, a variable stiffness compliance control method for the lower limb exoskeleton is designed based on the feedback linearization principle, which adjusts the stiffness of the lower limb exoskeleton in real time according to the online identified unmeasurable disturbance information and the control signal generated by the virtual control law to realize compliance control.

[0079] When designing the variable stiffness compliance control method for the lower limb exoskeleton based on the finite time disturbance prediction, the feedback linearization principle is used to convert the nonlinear dynamics model of the lower limb exoskeleton into a linear model, and the output of the virtual control law and the disturbance observer is combined to design a variable stiffness control strategy, which adjusts the stiffness of the lower limb exoskeleton in real time according to the disturbance prediction information. During the motion, the actual motion data of the lower limb exoskeleton, including joint angle, angular velocity, etc., are continuously obtained and input into the virtual control law and the finite time disturbance observer, and the control signal and the disturbance estimate are continuously updated, and then the stiffness is adjusted in real time to realize compliance control.

[0080] In certain embodiments, the variable stiffness control strategy includes combining the control signal generated by the virtual control law and the unmeasurable disturbance information identified by the finite time disturbance observer to determine a new control input, and adjusting the stiffness of the lower limb exoskeleton in real time according to the new control input and the joint state information of the lower limb exoskeleton. For example, the stiffness adjustment can be realized by changing the parameters of the elastic elements of the joint drive (such as the stiffness coefficient of the spring) or adjusting the stiffness coefficient in the control algorithm.

[0081] In some embodiments, based on the obtained virtual control law and the disturbance estimation result designed in step S4, according to the feedback linearization principle, a lower limb hip and knee joint sagittal plane motion dynamics input torque controller is constructed:

[0082]

[0083] Where k2 is the input torque controller gain.

[0084] The present application can improve the compliance of the hip and knee joint responses during the wearing of the lower limb exoskeleton, avoid the human-machine coordination consistency problems caused by the sudden increase in human-machine interaction force due to sudden changes in angle and speed during actual movement, eliminate the possible damage to the wearer's hip and knee joints, enhance the robustness of the lower limb exoskeleton wearable follow-up control, and realize real-time trajectory tracking of the lower limb exoskeleton hip and knee joints.

[0085] Exemplary Embodiments

[0086] Reference Figure 1 The lower limb exoskeleton model shown in the figure is a variable stiffness compliant control method for the lower limb exoskeleton based on finite time disturbance prediction. Figure 2 As shown, the following steps are included:

[0087] S1: Based on the Lagrangian modeling method, considering model uncertainty, mismatch, etc., a sagittal plane motion dynamics model of the lower limb hip and knee joints is established to characterize human-machine interaction forces and external disturbances:

[0088]

[0089] Where M0(q) represents the positive definite symmetric inertia matrix, is the centrifugal force and Coriolis force matrix, G0(q) represents the gravity matrix, q, Respectively represent the angle, angular velocity and angular acceleration of the lower limb exoskeleton hip and knee joints; ΔM(q), ΔG(q) is the uncertainty of model parameter identification, τ is the control torque input, τ h is the auxiliary human-machine interaction force, d is the external unmeasurable interference during the motion process, and its value is:

[0090] d=[2(sin(t)+sin(0.5t)),2(cos(0.5t)-cos(0.8t))] T (2)

[0091] Where t represents the time variable. d is a two-dimensional vector, with two components corresponding to the disturbances at the hip and knee joints, respectively. The hip joint disturbance d1 is the sum of sin(t) and sin(0.5t), simulating a combination of rapidly changing and slowly fluctuating disturbances. The knee joint disturbance d2 is the subtraction of cos(0.5t) and cos(0.8t), creating a periodic "beat frequency" effect. The disturbance changes with time t, reflecting dynamic uncertainty.

[0092] The angle q of the hip and knee joint is defined in scalar form as q = [q1, q2] T ; Among them, q1 is the angle of the hip joint, and q2 is the angle of the knee joint.

[0093] The identification result of the model parameter M0(q) is designed as:

[0094]

[0095] Where L1 = 0.436 m is the length of the thigh member of the system, L2 = 0.443 m is the length of the calf member of the system, m1 = 5.32 kg is the weight of the thigh member of the system, and m2 = 2.85 kg is the weight of the calf member of the system;

[0096] Model parameters The identification results are designed as:

[0097]

[0098] in, is the joint angular velocity;

[0099] The identification result of the model parameter G0(q) is designed as:

[0100]

[0101] Where g is the acceleration due to gravity, g = 9.8 m / s 2 ;

[0102] Simplify and merge the constructed dynamic models:

[0103]

[0104] in,

[0105] Define x1 = q and Design the sagittal kinematic / dynamic model of the lower limb hip and knee joints in state space form:

[0106]

[0107] in, represents the lumped disturbance of the lower limb exoskeleton control system including model parameter uncertainty and external disturbance, x1=[x 1,1 ,x 1,2 ] T ,x2=[x 2,1 ,x 2,2 ] T , x 1,1 Hip angle, x 12 Knee angle, x 21 Hip joint angular velocity, x 22 Knee joint angular velocity.

[0108] S2: Based on the dynamic relationship between human-machine interaction force and static balance force, an impedance model based on stiffness and damping is constructed to achieve the transformation from the desired angle trajectory to the reference angle trajectory:

[0109]

[0110] Among them, τ b represents the static equilibrium force, B and K represent the positive definite damping and stiffness matrices, the positive definite damping matrix is ​​B = diag{20,20}, the stiffness matrix is ​​K = diag{500,500}, x c represents the reference angle trajectory of the impedance model output, x d =[x d,1 ,x d,2 ] T Defined as the expected angle trajectory, the value is:

[0111]

[0112] τ b =[τ b,1 ,τ b,2 ] T represents the static equilibrium force, τ b，1 Hip joint and tau b，2 The knee joint values ​​are:

[0113]

[0114] Among them, x 1,i : actual joint angle (i = 1 hip, i = 2 knee), whose derivative represents the actual angular velocity; x c,i : The reference joint angle, whose derivative represents the desired angular velocity.

[0115] S3: Based on the reference angle trajectory design result (8), the virtual control law of the outer angle loop is designed using the feedback principle:

[0116]

[0117] Where k1=diag{2.04,2.33} is the control gain of the outer angle loop;

[0118] S4: Based on the constructed state-space kinematic / dynamic model of the lower limb hip and knee joint in the sagittal plane (7), and based on the finite-time convergence stability theorem, a finite-time disturbance observer is constructed to achieve online identification of unmeasurable disturbance information during motion:

[0119]

[0120] in, They represent the real-time angle, angular velocity and disturbance estimation results respectively. The observer design parameters are λ1=diag{7.3,6.9}, λ2=diag{7.6,7.5}, y=0.6, To design a function, its specific design is:

[0121]

[0122] S5: Based on the obtained virtual control law (11) and the disturbance estimation result (12) designed in step S4, according to the feedback linearization principle, the lower limb hip and knee joint sagittal plane motion dynamics input torque controller is constructed:

[0123]

[0124] Where k2=diag{4.07,4.66} is the input torque controller gain.

[0125] Through steps S1-S5, during the movement of wearing a lower limb exoskeleton, even when the external environmental interference presents non-periodic changes, by constructing a finite-time interference observer, it is possible to achieve online accurate estimation of the interference and compensate it in the controller, thereby enhancing the robustness of the lower limb exoskeleton when worn. By introducing an impedance model, it is possible to avoid the human-machine coordination consistency problem caused by the sudden increase in human-machine interaction force due to sudden changes in angle and speed in actual movement, thereby improving the compliance performance of the lower limb exoskeleton when worn.

[0126] In order to verify the effectiveness of the control method proposed in this application, simulation verification is carried out by taking the parameter setting and selection in this specific implementation as an example.

[0127] Figures 3-5 This is the simulation result. Figure 3 This is a schematic diagram of the trajectory tracking of the control method proposed in this application. Under the preset hip and knee joint angle reference angle trajectory, the control method proposed in this application can achieve accurate tracking of the reference angle. Figure 4This is a schematic diagram comparing the lumped interference estimate and the true value of the control method proposed in this application. It can be seen from the figure that the finite-time interference observer can well estimate the lumped interference, which provides an important basis for establishing a feedforward control law based on interference estimation and achieving accurate tracking of the time-varying desired angle trajectory signal. Figure 5 This is a schematic diagram of the angle trajectory adjustment under different stiffnesses of the control method proposed in this application. It can be seen from the figure that by setting the stiffness coefficient K of different impedance models, the reference angle trajectory can be adjusted to the desired angle trajectory to achieve the goal of smooth response.

[0128] The scope of protection claimed by this application is not limited to the above specific implementation methods. For those skilled in the art, this application may have various variations and changes. Any modifications, improvements and equivalent substitutions made within the concepts and principles of this application should be included in the scope of protection of this application.

Claims

1. A variable stiffness compliant control method for lower limb exoskeleton based on finite time disturbance prediction, characterized in that: include: Establish a dynamic model for the sagittal motion of the lower limb hip and knee joints that includes human-machine interaction forces and external disturbances; According to the dynamic relationship between human-machine interaction force and static balance force, an impedance model is designed based on the desired angle trajectory, and a reference angle trajectory under flexible target guidance is output through the impedance model; Constructing a virtual control law for an outer angle loop based on the reference angle trajectory and the current motion angle state, wherein the virtual control law is used to generate a control signal to guide the movement of the lower limb exoskeleton toward the reference angle trajectory; A finite-time disturbance observer is constructed based on the sagittal motion dynamics model of the hip and knee joints of the lower limb exoskeleton, and the unmeasurable disturbance information during the motion process is identified online using the finite-time disturbance observer; Utilizing the design results of the virtual control law and the finite-time disturbance observer, a variable stiffness compliant control method for the lower limb exoskeleton is designed based on the feedback linearization principle. This method adjusts the stiffness of the lower limb exoskeleton in real time according to the unmeasurable disturbance information identified online and the control signal generated by the virtual control law to achieve compliant control.

2. The control method according to claim 1, characterized in that: Based on the Lagrangian modeling method, a sagittal motion dynamics model of the lower limb hip and knee joints is established to characterize the human-machine interaction force and external disturbance: Where M0(q) represents the positive definite symmetric inertia matrix, is the centrifugal force and Coriolis force matrix, G0(q) represents the gravity matrix, q, Respectively represent the angle, angular velocity and angular acceleration of the hip and knee joints of the lower limb exoskeleton; ΔM(q), ΔG(q) is the uncertainty of model parameter identification, τ is the control torque input, τ h is the auxiliary human-computer interaction force, d is the external unmeasurable interference during the movement; Simplify and merge the constructed dynamic models: in, It represents the lumped disturbance of the lower limb exoskeleton control system including model parameter uncertainty and external disturbance; Define x1 = q and Design the sagittal kinematic / dynamic model of the lower limb hip and knee joints in state space form:

3. The control method according to claim 2, characterized in that: According to the dynamic relationship between human-machine interaction force and static balance force, an impedance model based on stiffness and damping is constructed to output the reference angle trajectory under flexible target guidance: Among them, τ b represents the static equilibrium force, B and K represent the positive definite damping and stiffness matrices respectively, x d Defined as the desired angle trajectory, x c Represents the reference angle trajectory of the impedance model output.

4. The control method according to claim 2, characterized in that: Based on the reference angle trajectory, the virtual control law of the outer angle loop is designed using the feedback principle: Where k1 is the control gain of the outer angle loop.

5. The control method according to claim 2, characterized in that: According to the sagittal plane kinematics / dynamics model of the lower limb hip and knee joints, the finite-time disturbance observer is constructed based on the finite-time convergence stability theorem: in, Represent the real-time angle, angular velocity and interference estimation results, λ1=diag{λ 1,1 ,λ 1,2 },λ2=diag{λ 2,1 ,λ 2,2 }, is the observer design parameter, 0<y<1, To design a function, its specific form is designed as follows:

6. The control method according to claim 2, characterized in that: Using the design results of the virtual control law and disturbance observer, according to the feedback linearization principle, the dynamic input torque controller of the sagittal plane motion of the lower limb hip and knee joints is constructed: Where k2 is the input torque controller gain.

7. A lower limb exoskeleton control device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, the method for controlling the variable stiffness and compliance of the lower limb exoskeleton according to any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium, characterized in that A computer program is stored, and when the computer program is executed by a processor, the variable stiffness and compliant control method for a lower limb exoskeleton according to any one of claims 1 to 6 is implemented.

Images