Fixed time backstepping trajectory tracking control method for mechanical arm with model uncertainty and external disturbance

By compensating the model uncertainty and external perturbation of the robot arm through adaptive radial basis function neural network, a fixed-time inverse step trajectory tracking controller is designed, which solves the problems of poor stability and low robustness of fixed-time control methods in the prior art. , realizes system convergence and performance improvement in fixed time.

CN118906048AActive Publication Date: 2024-11-08江淮前沿技术协同创新中心 +1

Patent Information

Application Number
CN202410973171.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-08
Estimated Expiration
2044-07-19

AI Technical Summary

Technical Problem

The existing fixed-time control method is poor in stability and low in robustness when facing model uncertainty and external disturbances of the robotic arm, and cannot meet the demand for rapid and stableness.

Method used

A fixed-time inverse step trajectory tracking control method based on adaptive radial basis function neural network is designed. By establishing a dynamic model of a multi-degree of freedom robot arm and a lumped perturbation estimation of external interference, fixed-time inverse step trajectory is designed. The tracking controller uses the adaptive radial basis function neural network to compensate the uncertainty and external perturbation of the model to achieve the convergence of the system within a fixed time range.

Benefits of technology

System convergence is achieved in a fixed time range independent of the initial state, improving the stability and robustness of the multi-degree of robotic arms when facing model uncertainty and external interference, and improving dynamic and steady-state performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118906048A_ABST
    Figure CN118906048A_ABST
Patent Text Reader

Abstract

The invention discloses a fixed-time backstepping trajectory tracking control method for a mechanical arm with model uncertainty and external disturbance, and belongs to the technical field of trajectory tracking control of a mechanical arm system. According to the invention, the problems of poor stability and low robustness of the existing fixed time control method are solved. The method comprises the following steps: firstly, establishing a dynamic model of a multi-degree-of-freedom mechanical arm in a joint space, wherein the dynamic model considers model uncertainty and external interference; then, approximation is conducted on the mechanical arm model uncertainty and the lumped disturbance of the external disturbance through the designed self-adaptive radial basis function neural network, and the estimated value of the model uncertainty and the estimated value of the lumped disturbance of the external disturbance are obtained; and finally, designing a fixed time backstepping trajectory tracking controller according to the established kinetic model and the lumped disturbance estimated value. The method can be used for controlling the mechanical arm with model uncertainty and external disturbance.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of trajectory tracking control of a robot system, and in particular relates to a fixed-time backstepping trajectory tracking control method of a robot with model uncertainty and external disturbance. Background Art

[0002] In recent years, robot trajectory tracking control has been widely used in industrial automation, robot operation, medical equipment, scientific research and other fields. These application backgrounds require the robot to accurately and stably track the given trajectory, improve production efficiency, reduce labor costs, reduce operational risks, and help achieve more precise experiments and operations. Considering the nonlinearity of the system and the existence of complex external interference, it is difficult to obtain the accurate model and environmental parameters of the controlled object.

[0003] Backstepping control is a nonlinear control method that aims to design a controller to achieve system stability and performance. It can handle complex nonlinear systems and offset system uncertainties and external disturbances through nonlinear compensators to improve the robustness and stability of the system.

[0004] However, the traditional backstepping control method can only ensure that the system achieves asymptotic convergence and cannot meet the requirements of rapid stability of the actual system. Finite-time backstepping control can ensure that the system achieves convergence within a finite time, but its convergence time is limited by the initial state of the system, and the actual application effect is not ideal. Therefore, fixed-time control has attracted the attention of researchers in recent years. The phenomenon of fixed-time convergence was first discovered by Andrieu, and its upper bound of convergence time is independent of the initial state of the system. Polyakov designed a fixed-time stable controller for uncertain linear objects, achieving robustness to matching uncertainty and external disturbances. Subsequently, researchers proposed a variety of fixed-time control algorithms based on backstepping control, but did not consider the convergence speed of the system too much, and did not consider the uncertainty of the model.

[0005] In summary, since the existing fixed-time control method does not consider the influence of model uncertainty, the stability and robustness of the existing fixed-time control method are still poor. It is very necessary to propose a fixed-time control method to solve the above problems. Summary of the invention

[0006] The purpose of the present invention is to solve the problems of poor stability and low robustness of existing fixed-time control methods, and to propose a fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbances.

[0007] The technical solution adopted by the present invention to solve the above technical problems is: a fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbances, the method specifically comprising the following steps:

[0008] Step 1: Establish a dynamic model of a multi-degree-of-freedom manipulator in joint space taking into account model uncertainty and external interference;

[0009] Step 2: Design an adaptive radial basis function neural network, use the adaptive radial basis function neural network to approximate the lumped disturbance of the robot model uncertainty and external interference, and obtain the lumped disturbance estimation value of the model uncertainty and external interference;

[0010] Step 3: Design a fixed-time backstepping trajectory tracking controller based on the dynamic model established in step 1 and the lumped disturbance estimate in step 2.

[0011] Furthermore, the dynamic model of the multi-degree-of-freedom manipulator in the joint space considering model uncertainty and external disturbance is:

[0012]

[0013] In the formula, q represents the vector composed of the joint angles of the robot arm, Represents the vector composed of the angular velocity of each joint of the robot arm, represents the vector composed of the angular acceleration of each joint of the robot arm, τ(t) is the vector composed of the control input torque of each joint, F(t) represents the viscous friction torque, represents a real number, n represents the number of degrees of freedom of the robot, G(q) is the gravity matrix, M(q) is the inertia matrix of the robot system, is the centrifugal force and Coriolis force matrix;

[0014] M(q)=M0(q)+ΔM(q) (2)

[0015]

[0016] Where M0(q) and C0(q) are the nominal model parameter matrices, ΔM(q) and are M0(q) and The uncertain part of

[0017] Consider the second-order system of the robot arm joint angle: x1 = q, The kinetic model of formula (1) is transformed into:

[0018]

[0019] In the formula, M0(q) is abbreviated as M0, the superscript -1 represents the inverse of the matrix, ΔM(q) is abbreviated as ΔM, and abbreviated as C, G(q) as G, F(t) as F, τ(t) as τ, is the first derivative of x1, is the first derivative of x2;

[0020] Define the second-order system of robot trajectory tracking error:

[0021] e1=x1-q r , e2=x2-α (5)

[0022] Where e1 and e2 are the error variables of the multi-degree-of-freedom manipulator, q r is the joint position reference signal, α is the vector composed of the backstepping auxiliary variables corresponding to each joint;

[0023] and They are:

[0024]

[0025] in, In order to consider the uncertainty of the multi-degree-of-freedom robot model and the lumped disturbance of external interference, is the first-order derivative of e1, is the first derivative of e2, is the joint velocity reference signal.

[0026] Furthermore, the specific form of the adaptive radial basis function neural network is:

[0027]

[0028] In the formula, δ(Z i ) is the output of the adaptive radial basis function neural network, that is, the model uncertainty of the i-th joint and the lumped disturbance estimate of the i-th joint subjected to external interference, δ = [δ(Z1),δ(Z2),…,δ(Z n )],ω i * is the weight matrix of the ith joint, is the radial basis function, Z=[z1,z2,z3] T is the input variable of the adaptive radial basis function neural network, z1=e1, z2=e2, z3=α, e1, e2, α are all n×1 dimensional vectors, Z i =[e 1i ,e 2i ,α i ]T ∈R 3×1 , R is a real number, e 1i ,e 2i ,α i are the i-th element in e1, e2, α, ε i is the estimation error of the ith joint, and for any small positive constant ε m ,|ε i |<ε m , |·| means taking the absolute value.

[0029] Furthermore, the radial basis function The jth component in for:

[0030]

[0031] In the formula, μ j is the center vector of the jth hidden layer node, b j is the width vector of the jth hidden layer node, j = 1, 2, …, k, k is the total number of hidden layer nodes, and the superscript T represents the transpose of the matrix.

[0032] Furthermore, the backstepping auxiliary variable α corresponding to the i-th joint i for:

[0033]

[0034] In the formula, k 1i and k 2i is the power coefficient of the backstepping auxiliary variable corresponding to the i-th joint, l1 and l2 are the proportional coefficients of the backstepping auxiliary variable, and l1>0, l2>0, and is the sign power transformation function.

[0035] Furthermore, the power coefficient k of the backstepping auxiliary variable corresponding to the i-th joint is 1i and k 2i for:

[0036]

[0037] Where m1 and m2 are switching constants, and m1>1,

[0038] Furthermore, the specific process of step three is:

[0039]

[0040] In the formula, τ iis the fixed-time backstepping trajectory tracking controller of the i-th joint, λ is a constant greater than zero, l3 and l4 are the proportional coefficients of the controller, and l3>0, l4>0, and is the sign power transformation function, k 3i and k 4i is the power coefficient of the controller corresponding to the i-th joint, a=diag[a1,a2,...,a n ],a1,a2,...,a n are all constants greater than 0. is θ i *T The estimated value of θ i *T is θ i * The transpose of i * is the weight vector of the i-th joint.

[0041] Furthermore, the weight vector θ of the i-th joint i * for:

[0042]

[0043] In the formula, ||ω i * || 2 is the i-th joint weight matrix ω i * The square of .

[0044] Furthermore, the power coefficient k of the controller corresponding to the i-th joint is 3i and k 4i for:

[0045]

[0046] Furthermore, the estimated value for:

[0047]

[0048] In the formula, is the adaptive law of the i-th joint weight, yes The transpose of yes The first derivative of , Γ i is the scale parameter of the adaptive law for the i-th joint weight, Γ i >0,σ 1i and σ 2i is the weight parameter of the adaptive law, σ1i >0,σ 2i >0,e 2i is the velocity error of the i-th joint of the robot, that is, e 2i is the i-th element in e2.

[0049] The beneficial effects of the present invention are:

[0050] The present invention designs an adaptive radial basis function neural network to observe the lumped disturbances of system model uncertainty and external interference, and designs an adaptive backstepping trajectory tracking controller based on the lumped disturbance estimation results, so that the system can converge within a fixed time range independent of the initial state, while improving the stability and robustness of the multi-degree-of-freedom robotic arm system in the face of model uncertainty and external interference.

[0051] The simulation results show that the method of the present invention can make the output trajectory of the robot system have better dynamic performance and steady-state performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of a fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance of the present invention;

[0053] Figure 2a It is the tracking trajectory waveform 1 under the traditional fixed-time adaptive backstepping control;

[0054] Figure 2b It is the tracking trajectory waveform 2 under the traditional fixed-time adaptive backstepping control;

[0055] Figure 3a It is the tracking error waveform 1 under the traditional fixed-time adaptive backstepping control;

[0056] Figure 3b It is the tracking error waveform 2 under the traditional fixed-time adaptive backstepping control;

[0057] Figure 4a It is the control torque waveform 1 under the traditional fixed-time adaptive backstepping control;

[0058] Figure 4b It is the second control torque waveform under the traditional fixed-time adaptive backstepping control;

[0059] Figure 5a It is the tracking trajectory waveform 1 under fast fixed-time adaptive backstepping control;

[0060] Figure 5b It is the tracking trajectory waveform 2 under fast fixed-time adaptive backstepping control;

[0061] Figure 6aIt is the tracking error waveform 1 under fast fixed time adaptive backstepping control;

[0062] Figure 6b It is the tracking error waveform 2 under fast fixed time adaptive backstepping control;

[0063] Figure 7a It is the tracking torque waveform 1 under fast fixed time adaptive backstepping control;

[0064] Figure 7b It is the tracking torque waveform 2 under fast fixed time adaptive backstepping control;

[0065] Figure 8a A tracking trajectory waveform 1 under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance of the present invention;

[0066] Figure 8b A tracking trajectory waveform 2 under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance of the present invention;

[0067] Figure 9a A tracking error waveform 1 under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance of the present invention;

[0068] Figure 9b A tracking error waveform 2 under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance of the present invention;

[0069] Fig.10a A control torque waveform 1 under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance of the present invention;

[0070] Fig.10b The invention discloses a second control torque waveform under fixed-time backstepping trajectory tracking control of a manipulator system with model uncertainty and external disturbance. DETAILED DESCRIPTION

[0071] Specific implementation method 1: Combination Figure 1 The present embodiment is described in detail. The present embodiment describes a fixed-time backstepping trajectory tracking control method for a robotic arm with model uncertainty and external disturbances, and the method specifically comprises the following steps:

[0072] Step 1: Establish a dynamic model of the multi-degree-of-freedom (the number of degrees of freedom is greater than or equal to 2) manipulator in the joint space taking into account model uncertainty and external interference;

[0073] Step 2: Design an adaptive radial basis function neural network, use the adaptive radial basis function neural network to approximate the lumped disturbance of the robot model uncertainty and external interference, and obtain the lumped disturbance estimation value of the model uncertainty and external interference;

[0074] Step 3: Design a fixed-time backstepping trajectory tracking controller based on the dynamic model established in step 1 and the lumped disturbance estimate in step 2.

[0075] This embodiment proposes a new fixed-time backstepping controller based on an adaptive neural network for the trajectory tracking problem of a robotic arm system with uncertainty. An adaptive radial basis function neural network is designed to compensate for the uncertainty of the model and external disturbances, and corresponding backstepping intermediate variables are designed to achieve rapid convergence of the controlled system within a fixed time range, ensuring that the multi-degree-of-freedom robotic arm system with uncertainty can accurately track the target output trajectory. The system can achieve convergence within a fixed time that is independent of the initial state, and the stability of the system is proved by using the Lyapunov stability theorem.

[0076] Specific implementation method 2: This implementation method is different from the specific implementation method 1 in that the dynamic model of the multi-degree-of-freedom manipulator in the joint space considering model uncertainty and external disturbance is:

[0077]

[0078] In the formula, q represents the vector composed of the joint angles of the robot arm, Represents the vector composed of the angular velocity of each joint of the robot arm, represents the vector composed of the angular acceleration of each joint of the robot arm, τ(t) is the vector composed of the control input torque of each joint, F(t) represents the viscous friction torque, represents a real number, n represents the number of degrees of freedom of the robot, G(q) is the gravity matrix, M(q) is the inertia matrix of the robot system, is the centrifugal force and Coriolis force matrix;

[0079] M(q)=M0(q)+ΔM(q) (2)

[0080]

[0081] Where M0(q) and C0(q) are the nominal model parameter matrices, ΔM(q) and are M0(q) and The uncertain part of

[0082] Consider the second-order system of the robot arm joint angle: x1 = q, The kinetic model of formula (1) is transformed into:

[0083]

[0084] In the formula, M0(q) is abbreviated as M0, the superscript -1 represents the inverse of the matrix, ΔM(q) is abbreviated as ΔM, and abbreviated as C, G(q) as G, F(t) as F, τ(t) as τ, is the first derivative of x1, is the first derivative of x2;

[0085] Define the second-order system of robot trajectory tracking error:

[0086] e1=x1-q r , e2=x2-α (5)

[0087] Where e1 and e2 are the error variables of the multi-degree-of-freedom manipulator, q r is the joint position reference signal, α is the vector composed of the backstepping auxiliary variables corresponding to each joint;

[0088] and They are:

[0089]

[0090] in, In order to consider the uncertainty of the multi-degree-of-freedom robot model and the lumped disturbance of external interference, is the first-order derivative of e1, is the first derivative of e2, is the joint velocity reference signal.

[0091] The other steps and parameters are the same as those in the first embodiment.

[0092] Specific implementation method three: This implementation method is different from specific implementation method one or two in that the specific form of the adaptive radial basis function neural network is:

[0093]

[0094] In the formula, δ(Z i ) is the output of the adaptive radial basis function neural network, that is, the model uncertainty of the i-th joint and the lumped disturbance estimate of the i-th joint subjected to external interference, δ = [δ(Z1),δ(Z2),…,δ(Z n )],ω i * is the weight matrix of the ith joint, is the radial basis function (the Gaussian radial basis function is used), Z=[z1,z2,z3] T is the input variable of the adaptive radial basis function neural network, z1=e1, z2=e2, z3=α, e1, e2, α are all n×1 dimensional vectors, Z i =[e 1i ,e 2i ,α i ] T ∈R 3×1 , R is a real number, e 1i ,e 2i ,α i are the i-th element in e1, e2, α, ε i is the estimation error of the ith joint, and for any small positive constant ε m ,|ε i |<ε m , |·| means taking the absolute value.

[0095] The other steps and parameters are the same as those in the first or second embodiment.

[0096] Specific implementation method 4: This implementation method is different from any one of the specific implementation methods 1 to 3 in that the radial basis function The jth component in for:

[0097]

[0098] In the formula, μ j is the center vector of the jth hidden layer node, the center vector μ j is a 3×1 column vector, b j is the width vector of the jth hidden layer node, j = 1, 2, …, k, k is the total number of hidden layer nodes, and the superscript T represents the transpose of the matrix.

[0099] The other steps and parameters are the same as those in Specific Embodiments 1 to 3.

[0100] Specific implementation method 5: This implementation method is different from the specific implementation methods 1 to 4 in that the backstepping auxiliary variable α corresponding to the i-th joint i for:

[0101]

[0102] In the formula, k 1i and k 2i is the power coefficient of the backstepping auxiliary variable corresponding to the i-th joint, l1 and l2 are the proportional coefficients of the backstepping auxiliary variable, and l1>0, l2>0, and is the sign power transformation function.

[0103] The other steps and parameters are the same as those in Specific Embodiments 1 to 4.

[0104] The symbolic power transformation function satisfies:

[0105] sig c x=[x1| c sign(x1),|x2| c sign(x2),...,|x m | c sign(x m )] T (10)

[0106] Here, sign(·) is the sign function, and x1 is the first element in x.

[0107] Specific implementation method 6: This implementation method is different from any one of the specific implementation methods 1 to 5 in that the power coefficient k of the backstepping auxiliary variable corresponding to the i-th joint is 1i and k 2i for:

[0108]

[0109] Where m1 and m2 are switching constants, and m1>1,

[0110] The other steps and parameters are the same as those in Specific Implementation Methods 1 to 5.

[0111] Specific implementation method 7: This implementation method is different from the specific implementation methods 1 to 6 in that the specific process of step 3 is as follows:

[0112]

[0113] In the formula, τ i is the fixed-time backstepping trajectory tracking controller of the i-th joint, λ is a constant greater than zero, l3 and l4 are the proportional coefficients of the controller, and l3>0, l4>0, and is the sign power transformation function, k 3i and k 4i is the power coefficient of the controller corresponding to the i-th joint, a=diag[a1,a2,…,a n ],a1,a2,…,a n are all constants greater than 0. is θ i *T The estimated value of θ i *T is θ i *The transpose of i * is the weight vector of the i-th joint.

[0114] The other steps and parameters are the same as those in Specific Embodiments 1 to 6.

[0115] In this implementation, the system output trajectory q is tracked to the given value q r A fixed-time backstepping trajectory tracking controller is proposed with the control objective of

[0116] Specific implementation eight: This implementation differs from any one of specific implementations one to seven in that the weight vector θ of the i-th joint i * for:

[0117]

[0118] In the formula, ||ω i * || 2 is the i-th joint weight matrix ω i * The square of .

[0119] The other steps and parameters are the same as those in Specific Embodiments 1 to 7.

[0120] Specific implementation method 9: This implementation method is different from any one of the specific implementation methods 1 to 8 in that the power coefficient k of the controller corresponding to the i-th joint is 3i and k 4i for:

[0121]

[0122] The other steps and parameters are the same as those in Specific Embodiments 1 to 8.

[0123] Specific implementation method 10: This implementation method is different from the specific implementation methods 1 to 9 in that the estimated value for:

[0124]

[0125] In the formula, is the adaptive law of the i-th joint weight, yes The transpose of yes The first derivative of , Γ i is the scale parameter of the adaptive law for the i-th joint weight, Γ i >0,σ 1i and σ 2i is the weight parameter of the adaptive law, σ1i >0,σ 2i >0,e 2i is the velocity error of the i-th joint of the robot, that is, e 2i is the i-th element in e2.

[0126] The other steps and parameters are the same as those in Specific Embodiments 1 to 9.

[0127] The effectiveness of the present invention is verified by simulation results. For a two-degree-of-freedom robotic arm system, the reference trajectory is selected as q r =[sin(0.5t)cos(0.5t)-1] T , the initial state is q0 = [-3 -2] T In order to prove the effectiveness of the present invention, the performance of the traditional fixed-time adaptive backstepping control method (i.e., method 1), the fast fixed-time adaptive backstepping control method (i.e., method 2) and the fixed-time backstepping control method proposed by the present invention are analyzed. The specific control parameters are shown in Table 1:

[0128] Table 1

[0129]

[0130] The controller of the traditional fixed-time adaptive backstepping control method is as follows:

[0131]

[0132] The controller of the fast fixed-time adaptive backstepping control method is as follows:

[0133]

[0134] Figure 2a , Figure 2b , Figure 5a , Figure 5b , Figure 8a as well as Figure 8b The tracking trajectory waveforms of the two-degree-of-freedom robot arm under three methods are shown. Figure 3a , Figure 3b , Figure 6a , Figure 6b , Figure 9a as well as Figure 9b The tracking error waveforms of the two-degree-of-freedom manipulator under three methods are shown. Figure 4a , Figure 4b , Figure 7a , Figure 7b , Fig.10a as well as Fig.10bThe control torque waveforms of the two-degree-of-freedom manipulator under the three methods are shown. From the results, it can be seen that the new fast backstepping fixed-time control method based on radial basis function neural network proposed in this invention achieves faster convergence speed and better convergence accuracy, and its convergence error is within 4×10 -3 rad, which verifies the effectiveness of the control algorithm designed by the present invention.

[0135] The above calculation examples of the present invention are only used to explain the calculation model and calculation process of the present invention in detail, and are not intended to limit the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbances, characterized in that: The method specifically comprises the following steps: Step 1: Establish a dynamic model of a multi-degree-of-freedom manipulator in joint space taking into account model uncertainty and external interference; Step 2: Design an adaptive radial basis function neural network, use the adaptive radial basis function neural network to approximate the lumped disturbance of the robot model uncertainty and external interference, and obtain the lumped disturbance estimation value of the model uncertainty and external interference; Step 3: Design a fixed-time backstepping trajectory tracking controller based on the dynamic model established in step 1 and the lumped disturbance estimate in step 2.

2. A fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbance according to claim 1, characterized in that: The dynamic model of the multi-degree-of-freedom manipulator in the joint space considering model uncertainty and external disturbance is: In the formula, q represents the vector composed of the joint angles of the robot arm, Represents the vector composed of the angular velocity of each joint of the robot arm, represents the vector composed of the angular acceleration of each joint of the robot arm, τ(t) is the vector composed of the control input torque of each joint, F(t) represents the viscous friction torque, represents a real number, n represents the number of degrees of freedom of the robot, G(q) is the gravity matrix, M(q) is the inertia matrix of the robot system, is the centrifugal force and Coriolis force matrix; M(q)=M0(q)+ΔM(q) (2) Where M0(q) and C0(q) are the nominal model parameter matrices, ΔM(q) and are M0(q) and The uncertain part of Consider the second-order system of the robot arm joint angle: x1 = q, The kinetic model of formula (1) is transformed into: In the formula, M0(q) is abbreviated as M0, the superscript -1 represents the inverse of the matrix, ΔM(q) is abbreviated as ΔM, and abbreviated as C, G(q) as G, F(t) as F, τ(t) as τ, is the first derivative of x1, is the first derivative of x2; Define the second-order system of robot trajectory tracking error: e1=x1-q r , e2=x2-α (5) Where e1 and e2 are the error variables of the multi-degree-of-freedom manipulator, q r is the joint position reference signal, α is the vector composed of the backstepping auxiliary variables corresponding to each joint; and They are: in, In order to consider the uncertainty of the multi-degree-of-freedom robot model and the lumped disturbance of external interference, is the first-order derivative of e1, is the first derivative of e2, is the joint velocity reference signal.

3. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 2, characterized in that: The specific form of the adaptive radial basis function neural network is: In the formula, δ(Z i ) is the output of the adaptive radial basis function neural network, that is, the model uncertainty of the i-th joint and the lumped disturbance estimate of the i-th joint subjected to external interference, δ = [δ(Z1),δ(Z2),…,δ(Z n )],ω i * is the weight matrix of the ith joint, is the radial basis function, Z=[z1,z2,z3] T is the input variable of the adaptive radial basis function neural network, z1=e1, z2=e2, z3=α, e1, e2, α are all n×1 dimensional vectors, Z i =[e 1i ,e 2i ,α i ] T ∈R 3×1 , R is a real number, e 1i ,e 2i ,α i are the i-th element in e1, e2, α, ε i is the estimation error of the ith joint, and for any small positive constant ε m ,|ε i |<ε m , |·| means taking the absolute value.

4. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 3, characterized in that: The radial basis function The jth component in for: In the formula, μ j is the center vector of the jth hidden layer node, b j is the width vector of the jth hidden layer node, j = 1, 2, …, k, k is the total number of hidden layer nodes, and the superscript T represents the transpose of the matrix.

5. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 3, characterized in that: The backstepping auxiliary variable α corresponding to the i-th joint i for: In the formula, k 1i and k 2i is the power coefficient of the backstepping auxiliary variable corresponding to the i-th joint, l1 and l2 are the proportional coefficients of the backstepping auxiliary variable, and l1>0, l2>0, and is the sign power transformation function.

6. A fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbance according to claim 5, characterized in that: The power coefficient k of the backstepping auxiliary variable corresponding to the i-th joint 1i and k 2i for: Where m1 and m2 are switching constants, and m1>1, 7. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 4, characterized in that: The specific process of step three is: In the formula, τ i is the fixed-time backstepping trajectory tracking controller of the i-th joint, λ is a constant greater than zero, l3 and l4 are the proportional coefficients of the controller, and l3>0, l4>0, and is the sign power transformation function, k 3i and k 4i is the power coefficient of the controller corresponding to the i-th joint, a=diag[a1,a2,...,a n ],a1,a2,...,a n are all constants greater than 0. is θ i *T The estimated value of θ i *T is θ i * The transpose of i * is the weight vector of the i-th joint.

8. A fixed-time backstepping trajectory tracking control method for a manipulator with model uncertainty and external disturbance according to claim 7, characterized in that: The weight vector θ of the i-th joint i * for: In the formula, ||ω i * || 2 is the i-th joint weight matrix ω i * The square of .

9. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 7, characterized in that: The power coefficient k of the controller corresponding to the i-th joint 3i and k 4i for:

10. The fixed-time backstepping trajectory tracking control method of a manipulator with model uncertainty and external disturbance according to claim 7, characterized in that: The estimated value for: In the formula, is the adaptive law of the i-th joint weight, yes The transpose of yes The first derivative of , Γ i is the scale parameter of the adaptive law for the i-th joint weight, Γ i >0,σ 1i and σ 2i is the weight parameter of the adaptive law, σ 1i >0,σ 2i >0,e 2i is the velocity error of the i-th joint of the robot, that is, e 2i is the i-th element in e2.

Citation Information

Patent Citations

  • Finite time self-adaptive back stepping control method for flexible joint mechanical arm based on neural networks

    CN107662208A

  • Multi-joint industrial mechanical arm backstepping finite time sliding mode control method

    CN108803324A

  • Hybrid mechanism adaptive backstepping control method based on extreme learning machine

    CN110389525A

  • Neural-network-based adaptive finite time command filtering backstepping control method

    CN110687796A

  • Uncertainty robot adaptive neural network control method based on inversion method

    CN112192573A

Cited By

  • Submarine crawler trajectory tracking control method based on adaptive backstepping control

    CN119882720A

  • Preset time point control method and system for limited state of mechanical arm

    CN121468555A