Control method of lower limb exoskeleton robot under saturated input and external disturbance

By establishing a lower limb exoskeleton dynamic model and adaptive controller, the problems of saturated input and external interference are solved, and the system stability and tracking performance are improved, ensuring the safety and adaptability of the lower limb exoskeleton robot.

CN117666344BActive Publication Date: 2024-11-05KUNMING UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311654022.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-05
Publication Date
2024-11-05
Estimated Expiration
2043-12-05

AI Technical Summary

Technical Problem

The existing lower limb exoskeleton robot control technology is difficult to effectively deal with saturated inputs and external interference, resulting in system out of control and safety risks, and it is difficult to achieve stable tracking and adaptability.

Method used

By establishing a lower limb exoskeleton dynamic model, determining the saturation input function, setting preset performance functions and Lyapunov functions, designing an adaptive controller to suppress external interference online, building a virtual controller and adaptive law, and achieving the stability and reference trajectory of the system track.

Benefits of technology

It improves the stability and tracking performance of the system, can effectively handle saturated inputs and external interference, ensures the safety and adaptability of the system, and achieves accurate tracking of the reference trajectory.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117666344B_ABST
    Figure CN117666344B_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of lower limb exoskeleton robot control, and discloses a lower limb exoskeleton robot control method under saturated input and external interference, establishes a dynamic model of the lower limb exoskeleton, then adopts a saturated approximation method to approximate the saturated input function, and sets an adaptive controller to improve the stability and performance of the system. The lower limb exoskeleton robot control method under saturated input and external interference considers the saturated input from the input of the actual exoskeleton through an adaptive controller. Different from other methods that do not fully consider the influence of saturated input, the invention converts the control problem under saturated input into the control problem of a nonlinear system with unknown control direction, and constructs an adaptive controller to suppress unknown external interference in real time online, which can ensure the stability of the system and realize the tracking of the reference trajectory.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of lower limb exoskeleton robot control, and in particular to a lower limb exoskeleton robot control method under saturated input and external interference. Background Art

[0002] With the continuous progress of modern medical science and technology, remarkable achievements have been made in the treatment of cerebrovascular and nervous system diseases such as stroke. However, even if they are lucky enough to receive treatment and survive, many patients still face serious sequelae such as hemiplegia, abnormal gait or paralysis, which makes their lives full of pain and suffering. In addition, it is worth noting that the average age of stroke patients is getting younger, and many young and middle-aged groups (20 to 54 years old) are also facing disability problems caused by muscle, bone and nervous system diseases. Although the current medical level has been greatly improved, there is currently a serious shortage of rehabilitation physicians in my country, which makes it difficult for many patients to obtain effective and timely rehabilitation treatment. However, with the continuous development of society and the rapid progress of science and technology, the field of medical rehabilitation is ushering in high-speed and high-quality development opportunities. In order to solve the current problem of shortage of rehabilitation medical resources in my country, a type of wearable exoskeleton for rehabilitation has begun to gradually replace traditional rehabilitation medical devices. Among them, human-computer interactive rehabilitation exoskeletons have been developed for rehabilitation training, such as upper limb rehabilitation exoskeletons and lower limb rehabilitation exoskeletons. The control problem of this type of wearable exoskeleton robot that can be used for rehabilitation has undoubtedly attracted widespread attention from academia and industry.

[0003] Among many rehabilitation exoskeletons, lower limb exoskeletons can provide important medical and rehabilitation support. Their superior performance, such as strong load-bearing capacity, high flexibility, energy-saving and high efficiency, safety and reliability, and high comfort, have attracted many scholars to study them. However, developing lower limb exoskeleton control with human-machine collaboration is a complex and arduous task, which involves solving multiple problems, including complex interactions, instantaneous decision-making, safety, adaptability, saturated input, and external interference.

[0004] In order to solve the control problem of lower limb exoskeleton, researchers have done a lot of work. For example, PID controllers are designed to implement some basic tracking tasks. However, it relies heavily on control parameters and is difficult to cope with external disturbances, especially the human-computer interaction between the exoskeleton and the wearer. In practice, the dynamic behavior of lower limb exoskeleton is very complex. It is not only difficult to establish an accurate mathematical model, but also encounters unknown interference from the outside world. In order to solve this complex control problem, a variety of adaptive control methods have been used to deal with external interference. Although some traditional or advanced adaptive methods, such as direct adaptation, adaptive robust control, reduced-order adaptation, etc., can indeed cope with external interference well, these methods basically do not take into account the situation of saturated input.

[0005] In exoskeleton systems, the importance of input saturation lies in that it can affect the performance and adaptability of the system. In addition, input saturation is directly related to the safety of the system. If the system cannot handle inputs beyond its design limits, it may cause the system to lose control or cause unpredictable harm to the wearer. Therefore, for lower limb exoskeletons, it is crucial to consider saturated inputs. At the same time, whether it is possible to solve the external interference that exists in the process of wearing the exoskeleton, especially the human-computer interaction force, is a crucial issue. Summary of the invention

[0006] 1. Technical issues to be solved

[0007] In view of the deficiencies in the prior art, the present invention provides a lower limb exoskeleton robot control method under saturated input and external interference, which has the advantages of improving the stability of the system and solves the above-mentioned technical problems.

[0008] (II) Technical solution

[0009] To achieve the above object, the present invention provides the following technical solution: a method for controlling a lower limb exoskeleton robot under saturated input and external interference, comprising the following steps:

[0010] S1. Establish a lower limb exoskeleton dynamic model;

[0011] S2, determining the saturation input function;

[0012] S3, setting a preset performance function, calculating the control law, the virtual controller and the adaptive law, and converging the model output;

[0013] S4. Analyze the output results.

[0014] As a preferred technical solution of the present invention, the system expression for establishing the lower limb exoskeleton dynamics model in step S1 is as follows:

[0015]

[0016] M=I r +I h

[0017] G=m r l r g+m s l s g

[0018] Among them, q, They represent the states of the hip joint, namely position, velocity and acceleration, M is the system inertia, I r represents the inertia of the exoskeleton system, I h represents the inertia of the human thigh, θ and ψ represent the viscosity and solid friction coefficients respectively, G is the gravitational term, m r 、m s are the thigh mass and the exoskeleton mass, respectively, l r , l s Represent the length of the thigh and the length of the exoskeleton, τ im represents the human-machine interaction force in the exoskeleton, sin(*) represents the internal sinusoidal operation, sign(*) represents the sign function, and τ(u) represents the input of the system.

[0019] As a preferred technical solution of the present invention, the process of determining the saturation input function in step S2 is as follows:

[0020] S2.1, setting the estimation function according to the saturation characteristics of the input;

[0021] S2.2, performing error approximation on the function of the saturation characteristic of the input by estimating the function;

[0022] S2.3. Define the function variables and calculate the new exoskeleton system equation.

[0023] As a preferred technical solution of the present invention, the model expression of the saturation characteristic u(v) input in step S2.1 is as follows:

[0024]

[0025] Among them, u M represents the upper limit of input saturation, u(τ) represents the control input of joint torque, v is the control variable, and sign(*) represents the sign function.

[0026] As a preferred technical solution of the present invention, the expression of the estimation function is as follows:

[0027]

[0028] u(v)=h(v)+Δ(v)

[0029] Where h(v) is the estimated function, tanh(*) represents the hyperbolic sine function, and Δ(v) represents the M The approximation error is bounded by [1-tanh(1)], and the function variable expression is defined as follows:

[0030]

[0031] x1=q

[0032]

[0033] Among them, x1 and x2 represent the system status, It means that the partial derivative of h(v) with respect to v is at v=v * The value of g(v * ) represents the control coefficient, v * The new input control, and v * ∈(0, v), the calculated exoskeleton system state space expression is as follows:

[0034]

[0035]

[0036]

[0037] J(v)=Δ(v)+h(0)

[0038] Among them, f(x1, x2) represents the input of the system, d is (x1, x2, v) represents the total interference term of the system, g(v * ) represents the control coefficient, J(v) represents the external disturbance and the disturbance term generated in the process of processing input saturation, and y represents the output of the system.

[0039] As a preferred technical solution of the present invention, the specific steps of step S3 are as follows:

[0040] S3.1. Setting preset performance functions;

[0041] S3.2. Define a variable Y that satisfies a preset performance function;

[0042] S3.3, set a Lyapunov function V;

[0043] S3.4. Establish the control law, virtual controller and adaptive law.

[0044] As a preferred technical solution of the present invention, the preset performance function expression in step S3.1 is as follows:

[0045]

[0046] Among them, ρ(t) represents the preset performance function, ρ0, σ represents three parameters that are all positive, and ρ0> e represents natural logarithm, and t represents time.

[0047] As a preferred technical solution of the present invention, the variable Y in the function in step S3.2 is expressed as follows:

[0048]

[0049] e1=qy d

[0050] Among them, e1 represents the tracking error of the system, q represents the joint angle, and y d represents a smooth and bounded reference signal, the variable Y and its corresponding initial value Y(0) satisfy the following expression:

[0051] 0<Y 2 <1

[0052] -ρ(0)<Y(0)<ρ(0)

[0053] Wherein, ρ(0) represents the initial value of the preset performance function.

[0054] As a preferred technical solution of the present invention, the expression of the Lyapunov function V in step S3.3 is as follows:

[0055]

[0056] in, represents the estimated error of the disturbance term, e2 represents the state error between the system controller α and the system, represents a parameter greater than 0, and ln(*) represents a logarithmic function.

[0057] As a preferred technical solution of the present invention, the expression of the control law can be obtained according to the Lyapunov function V as follows:

[0058]

[0059]

[0060] Among them, N(ζ) represents the Nussbaum function, k2 represents a parameter greater than 0, sgn represents that the internal value takes different positive and negative values ​​according to different conditions, and K represents a known variable. represents the first-order derivative of the system controller α, represents the estimated value of the interference term, and satisfy d is * represents the interference boundary, cos(*) represents the cosine function, f represents the modeled part of the system, Represents a parameter greater than 0, and the virtual controller and adaptive law expressions are as follows:

[0061]

[0062]

[0063]

[0064] in, express The first-order derivative of , α represents the system controller, Represents y d The first derivative of represents the first-order derivative of the preset performance function, k1 and μ represent parameters greater than 0, represents the first-order derivative of e2, represents the first-order derivative with respect to ζ.

[0065] Compared with the prior art, the present invention provides a control method for a lower limb exoskeleton robot under saturated input and external interference, which has the following beneficial effects:

[0066] The present invention transforms the control problem under saturated input into the control problem of a nonlinear system with unknown control direction, and constructs an adaptive controller to suppress unknown external interference online in real time, which can ensure the stability of the system and track the reference trajectory. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a schematic diagram of human-machine interaction of the exoskeleton of the present invention;

[0068] Figure 2 For the present invention * Trajectory diagram;

[0069] Figure 3 It is a schematic diagram of position tracking of the present invention;

[0070] Figure 4 A schematic diagram of tracking error of the present invention;

[0071] Figure 5 A schematic diagram comparing the position tracking of the PD control of the present invention and the control method proposed by the present invention;

[0072] Figure 6 A schematic diagram comparing the position tracking errors of the PD control of the present invention and the control method proposed by the present invention;

[0073] Figure 7 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0074] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0075] See also Figure 1-7 , a control method for a lower limb exoskeleton robot under saturated input and external disturbance, comprising the following steps:

[0076] S1. Establish a lower limb exoskeleton dynamics model. The system equation established according to the Newton-Euler formula is:

[0077]

[0078] Among them, q, They represent the states of the hip joint, i.e., position, velocity, and acceleration, respectively. M is the system inertia, and M = I r +I h , I r represents the inertia of the exoskeleton system, I h represents the inertia of the human thigh, θ and ψ represent the viscosity and solid friction coefficients respectively, G is the gravitational term, G = m r l r g+m s l s g,m r 、m s are the thigh mass and the exoskeleton mass, respectively, l r , l s Represent the length of the thigh and the length of the exoskeleton, τ im represents the human-machine interaction force in the exoskeleton, sin(*) represents the internal sinusoidal operation, sign(*) represents the sign function, and τ(u) is the input of the system;

[0079] S2. Determine the saturation input function. The saturation characteristics of the input can be expressed as:

[0080]

[0081] Among them, u(v) represents the estimation function, uM Indicates the upper limit of input saturation, u(τ) represents the control input of joint torque in HER, v is the control variable, sign(*) represents the sign function. For the input saturation of the system, how to deal with the piecewise problem of this saturation function is undoubtedly a major difficulty, which will bring great challenges to the design of the controller. In order to solve the problem of saturated input of the exoskeleton system, an estimation function h(v) is used to approximate the piecewise saturation function u(v), and the expression is as follows:

[0082]

[0083] u(v)=h(v)+Δ(v)

[0084] Where h(v) is the estimated function, tanh(*) represents the hyperbolic sine function, and Δ(v) represents the M The approximation error is bounded by [1-tanh(1)], and then the middle value theorem is used to estimate the function h(v):

[0085]

[0086] Then define the function variables, the expression is as follows:

[0087]

[0088] x1=q

[0089]

[0090] Among them, x1 and x2 represent the system status, It means that the partial derivative of h(v) with respect to v is at v=v * The value of g(v * ) represents the control coefficient, v * The new input control, and v * ∈(0, v), the calculated exoskeleton system state space expression is as follows:

[0091]

[0092] in, f(x1, x2) is the modeled part of the system, d is (x1, x2, v) represents the total interference term of the system, g(v * ) represents the control coefficient, J(v) represents the external interference and the interference term generated in the process of processing input saturation, J(v)=Δ(v)+h(0), the output, package and input problems of the y system will also be transformed into problems of unknown system control direction.

[0093] To this end, this project introduces a Nussbaum function to solve the problem of unknown control direction. Among them, the selectable Nussbaum type functions generally include the following: 2 cosζ,ζ 2 sinζ, Then, an adaptive controller is designed so that the output of the system can track the ideal reference trajectory. At the same time, the closed-loop system is eventually uniformly bounded (UUB) and the tracking error has a preset convergence performance, that is, e1 moves inward along the boundary of ρ(t).

[0094] S3. Then set the preset performance function:

[0095] ρ(t)=ρ0e -σt +ρ ò

[0096] Among them, ρ(t) represents the preset performance function, ρ0, σ represents three parameters that are all positive, and ρ0> e represents the natural logarithm, t represents the time, based on this, the tracking error of the system e1 = qy d , yd represents a smooth and bounded reference signal such that y d and its derivative Bounded, then define a variable that satisfies the preset performance function Ensure that When 0<Y 2 <1 is always true, and its initial value Y(0) should satisfy -ρ(0)<Y(0)<ρ(0)

[0097] Next, set a Lyapunov function V, the expression is as follows:

[0098]

[0099] in, represents the estimated error of the disturbance term, e2 represents the state error between the system controller α and the system, represents a parameter greater than 0, ln(*) represents a logarithmic function, and taking the derivative of V yields:

[0100]

[0101] From this, we can get the control law, the expression of which is as follows:

[0102]

[0103]

[0104] Among them, N(ζ) represents the Nussbaum function, k2 represents a parameter greater than 0, sgn represents that the internal value takes different positive and negative values ​​according to different conditions, and K represents a known variable. represents the first-order derivative of the system controller α, represents the estimated value of the interference term, and satisfy d is * represents the interference boundary, cos(*) represents the cosine function, f represents the modeled part of the system, Represents a parameter greater than 0, accompanied by a virtual controller and adaptive law as follows:

[0105]

[0106]

[0107]

[0108] in, express The first-order derivative of , α represents the system controller, represents the first-order derivative of yd, represents the first-order derivative of the preset performance function, k1 and μ represent parameters greater than 0, represents the first-order derivative of e2, represents the first-order derivative with respect to ζ;

[0109] S4, analyzing the output results;

[0110] like Figure 3 As shown in the figure, a position tracking simulation experiment using the method proposed in the present invention is given. It can be seen from the figure that the position curve output by the controller designed in the present invention can track the reference trajectory well and almost coincide with it. It can also be seen in the enlarged detail figure that the output of the system will converge within 0.4s. This shows that the method proposed in the present invention can ensure the stability and convergence of the system output, such as Figure 4 The graph of the position tracking error and the preset performance function is shown. It can be clearly seen in the figure that the tracking error moves inward along the PPF. The results show that the closed-loop system is eventually uniformly bounded (UUB) and the tracking error has the preset convergence performance, that is, e1 moves inward along the boundary of ρ(t) (that is, -ρ(t)<e1(t)<ρ(t));

[0111] from Figure 5 It can be seen that compared with the PD controller, by adopting the control method proposed in the present invention, the output trajectory of the system is closer to and coincides with the reference trajectory. Figure 6It can be seen that even if external disturbances are introduced, the control method proposed in the present invention makes the tracking error converge to a set of zeros asymptotically. However, PD control has a large error and does not converge. Obviously, the method proposed in the present invention has good tracking performance. In addition, the stability of the entire closed-loop system can also be guaranteed.

[0112] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for a lower limb exoskeleton robot under saturated input and external disturbance, characterized in that: The following steps are involved: S1. Establish a lower limb exoskeleton dynamic model; S2, determining the saturation input function; S3, setting a preset performance function, calculating the control law, the virtual controller and the adaptive law, and converging the model output; The specific steps of step S3 are as follows: S3.

1. Setting preset performance functions; The preset performance function expression in step S3.1 is as follows: Among them, ρ(t) represents the preset performance function, ρ0, σ represents three positive parameters, and e represents natural logarithm, t represents time; S3.

2. Define a variable Y that satisfies a preset performance function; The variable Y in the function in step S3.2 is expressed as follows: e16qy d Among them, e1 represents the tracking error of the system, q represents the joint angle, and y d represents a smooth and bounded reference signal, the variable Y and its corresponding initial value Y(0) satisfy the following expression: 0<Y 2 <1 -ρ(0) <Y(0)<ρ(0) Wherein, ρ(0) represents the initial value of the preset performance function; S3.3, set a Lyapunov function V; The expression of the Lyapunov function V in step S3.3 is as follows: in, represents the estimated error of the disturbance term, e2 represents the state error between the system controller α and the system, represents a parameter greater than 0, ln(*) represents a logarithmic function; S3.4, establishing the control law, virtual controller and adaptive law; According to the Lyapunov function V, the expression of the control law can be obtained as follows: Where N(ζ) represents the Nussbaum function, k2 represents a parameter greater than 0, sgn represents the positive and negative values ​​of the internal value according to different conditions, K represents a known variable and satisfies K=(Y / (ρ(t)(1-Y 2 ))), represents the first-order derivative of the system controller α, represents the estimated value of the interference term, and satisfy d is * represents the interference boundary, cos(*) represents the cosine function, f represents the modeled part of the system, Represents a parameter greater than 0, and the virtual controller and adaptive law expressions are as follows: in, express The first-order derivative of , α represents the system controller, represents the first-order derivative of yd, represents the first-order derivative of the preset performance function, k1 and μ represent parameters greater than 0, represents the first-order derivative of e2, represents the first-order derivative with respect to ζ; S4. Analyze the output results.

2. The method for controlling a lower limb exoskeleton robot under saturated input and external disturbance according to claim 1, characterized in that: The system expression for establishing the lower limb exoskeleton dynamics model in step S1 is as follows: M=I r +I h G=m r l r g+m s l s g Among them, q, They represent the states of the hip joint, namely position, velocity and acceleration, M is the system inertia, I r represents the inertia of the exoskeleton system, I h represents the inertia of the human thigh, θ and ψ represent the viscosity and solid friction coefficients respectively, G is the gravitational term, m r 、m s are the thigh mass and the exoskeleton mass, respectively, l r , l s Represent the length of the thigh and the length of the exoskeleton, τ im represents the human-machine interaction force in the exoskeleton, sin(*) represents the internal sinusoidal operation, sign(*) represents the sign function, and τ(u) represents the input of the system.

3. The method for controlling a lower limb exoskeleton robot under saturated input and external disturbance according to claim 2, characterized in that: The saturation input function determination process in step S2 is as follows: S2.1, setting the estimation function according to the saturation characteristics of the input; S2.2, performing error approximation on the function of the saturation characteristic of the input by estimating the function; S2.

3. Define the function variables and calculate the new exoskeleton system equation.

4. The method for controlling a lower limb exoskeleton robot under saturated input and external disturbance according to claim 3, characterized in that: The model expression of the saturation characteristic u(v) input in step S2.1 is as follows: Among them, u M represents the upper limit of input saturation, u(τ) represents the control input of joint torque, v is the control variable, and sign(*) represents the sign function.

5. The method for controlling a lower limb exoskeleton robot under saturated input and external disturbance according to claim 4, characterized in that: The expression of the estimation function is as follows: u(v)=h(v)+Δ(v) Among them, h(v) is the estimation function, tanh(*) represents the hyperbolic sine function, and Δ(v) represents the M The approximation error is bounded by [1-tanh(1)], and the function variable expression is defined as follows: x1=q Among them, x1 and x2 represent the system status, represents the value of the partial derivative of h(v) with respect to v when v = v*, g(v*) represents the control coefficient, v* is the new input control, and v*∈(0, v), the calculated state space expression of the exoskeleton system is as follows: J(v)=Δ(v)+h(0) Among them, f(x1, x2) represents the modeled part of the system, d is (x1, x2, v) represents the total interference term of the system, g(v * ) represents the control coefficient, J(v) represents the external disturbance and the disturbance term generated in the process of processing input saturation, and y represents the output of the system.