Humanoid robot finite time path tracking control method

Abstract

The invention discloses a humanoid robot finite time path tracking control method, and relates to the technical field of robot control, in particular to a humanoid robot finite time path tracking control realized by adopting an active-disturbance-rejection disturbance observer and a second-order sliding mode, and the specific steps are as follows: establishing a matrix expression model of humanoid robot nonlinear dynamics; designing an angular velocity subsystem control model; designing a humanoid robot displacement subsystem model; designing an active-disturbance-rejection disturbance observer; the control input vector of the second-order sliding mode finite time trajectory tracking is designed, so that the actual displacement of the humanoid robot can accurately track the given displacement, and the control parameters and the control input vector of the second-order sliding mode finite time trajectory tracking are adjusted, so that the tracking error of the humanoid robot completes minimum convergence within the finite time. According to the method, the external environment interference can be estimated within finite time, the system uncertainty caused by parameter perturbation can be effectively compensated, and the stability, robustness and adaptive capacity of the humanoid robot are improved.

Description

Technical Field

[0001] The invention relates to the technical field of robot control, and in particular to a finite-time path tracking control method for a humanoid robot. Background Art

[0002] By autonomously adjusting its position and posture, the humanoid robot can track three-dimensional trajectories and complete tasks such as fire rescue, resource exploration, power maintenance, and spacecraft capture. In order to accurately track the trajectory of humanoid robots under the combined effects of external disturbances and unmodeled dynamics, the commonly used methods for humanoid robot path tracking currently include force feedback control methods, including non-singular terminal sliding mode finite time controllers, fuzzy logic strategies, neural network technology, and adaptive terminal sliding mode capture controllers.

[0003] After searching relevant patents and documents at home and abroad, there are references such as "Application of Lyapunov-like theory in tracking humanoid robot task space" regarding the milling motion control method of thin-walled parts. It takes into account various uncertain factors of humanoid robots, proposes a Lyapunov-like method, and designs a task space controller with robustness. The controller obtained by this method not only achieves stable tracking within a limited deterministic time, but also does not require the inversion of the Jacobian matrix, and has robustness to certain types of external interference. However, this method is difficult to achieve a balance effect when ensuring smooth movement transition. There is also a document "Humanoid robot teleoperation motion redirection technology based on motion capture", which uses motion redirection technology based on error optimization to map human motion capture data to a humanoid robot to imitate human motion to complete teleoperation tasks. At the same time, it can automatically switch between the center of mass adjustment and the three-point support method to maintain balance during the teleoperation process, but this method does not have a significant inhibitory effect on the unstable oscillation behavior of the system. Another example is the invention patent with patent application number: CN202310103939.X, "A humanoid robot trajectory tracking method and trajectory tracking system", which collects the image to be processed, pre-processes the image, divides the image into five rectangular areas horizontally, and obtains the robot motion instructions according to the trajectory type and the coordinate position of the trajectory center point, judges the correctness of the motion instructions, and determines the final motion instructions of the robot. This method increases the difficulty of detection because of its high requirements on the accuracy of the mathematical model.

[0004] Although the above patents partially study the path tracking control methods of humanoid robots, the control methods are relatively complex, lack flexibility, and increase controller jitter. The present invention proposes a finite-time path tracking control method for a humanoid robot through an anti-disturbance observer and a second-order sliding mode, which can converge to the steady-state error bound in a shorter time, realize finite-time path tracking, and improve the stability and robustness of the humanoid robot. Summary of the invention

[0005] The present invention provides a finite-time path tracking control method for a humanoid robot, which solves the problem that the existing humanoid robot control method proposed in the above background technology is relatively complex, lacks flexibility, and increases controller jitter.

[0006] The present invention provides the following technical solution: a finite-time path tracking control method for a humanoid robot, which realizes a finite-time path tracking control of a humanoid robot by using an auto-disturbance rejection observer and a second-order sliding mode, comprising the following steps:

[0007] S1. The impedance control and control torque are analyzed by Newton-Euler equation, and the matrix expression model of nonlinear dynamics of humanoid robot is established. The established matrix expression model is:

[0008]

[0009] Where: J is the transformation matrix between the inertial coordinate system and the mechanism coordinate system, M(θ e ) is a real symmetric inertia matrix, is the Coriolis force and centrifugal force matrix, H(θ e ) is the gravity matrix, η is the attitude angular displacement, and g(η) is the restoring force matrix;

[0010] S2, dividing the humanoid robot kinematic speed controller into an angular velocity subsystem controller and a position subsystem controller;

[0011] S3. Design an angular velocity subsystem control model for the matrix expression model of the nonlinear dynamics of the humanoid robot, so that the actual angular velocity can converge to the desired angular velocity within a limited time; select an integral sliding surface for the angular velocity subsystem control model;

[0012] The angular velocity subsystem control model is:

[0013] Where: c is the actual angular velocity, is the expected angular velocity, α is a positive number, k is a constant, θ e is the angle;

[0014] S4. Based on the actual angular velocity obtained from the matrix expression model of the nonlinear dynamics of the humanoid robot, the displacement subsystem model of the humanoid robot is designed:

[0015]

[0016] Where: v c is the actual line speed, is the expected linear velocity, x e ande represent the displacement of the humanoid robot in the x and y directions respectively;

[0017] S5. Design an auto-disturbance rejection observer for the displacement subsystem model of the humanoid robot:

[0018]

[0019] Where: tanh(·) is a kind of hyperbolic function, sig(·) is the Gaussian radial basis function, λ is the constraint force vector, N and I are the mass and inertia matrix of the humanoid robot, δ is the external disturbance, is the observation error;

[0020] S6. Design a control input vector τ for second-order sliding mode finite-time trajectory tracking so that the actual displacement of the humanoid robot can accurately track the given displacement;

[0021]

[0022] Where: σ and ρ are the driving torques generated by the servo motors of the humanoid robot, Γ(·) is the gamma function, r1 and r2 are positive constants;

[0023] S7, adjusting the control parameters and the control input vector of the second-order sliding mode finite-time trajectory tracking so that the tracking error of the humanoid robot completes minimum convergence within a finite time.

[0024] Preferably, the Lyapunov function selected in S7 is:

[0025] Where: Derivative V(τ) is used to adjust the control parameters. When By using the control input vector τ of the second-order sliding mode finite-time trajectory tracking, the tracking error of the humanoid robot can be minimized within a finite time.

[0026] Preferably, for the matrix expression model of the nonlinear dynamics of the humanoid robot, the two variables α and k are adjusted, and the angular velocity subsystem control model is designed so that the actual angular velocity can converge to the desired angular velocity within a limited time.

[0027] Preferably, the selected integral sliding surface is:

[0028]

[0029] Where ξ is an auxiliary variable, λ1 and λ2 are both positive numbers, and tan(·) is the tangent function.

[0030] Preferably, when the ADRO accurately estimates the external disturbance, the observation error Converges to a neighborhood near the origin in finite time.

[0031] Preferably, an anti-disturbance observer and a second-order sliding mode are used to realize finite-time path tracking control of a humanoid robot, compensate for system uncertainty caused by parameter perturbations, have the ability to online adjust weights and control parameters of the humanoid robot, and have the characteristics of a simple motion model and weak coupling of degrees of freedom.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. The finite-time path tracking control method of the humanoid robot adopts an anti-disturbance observer and a second-order sliding mode to realize the finite-time path tracking control of the humanoid robot. It can estimate the external environmental interference within a finite time, effectively compensate for the system uncertainty caused by parameter perturbations, and has a better fault-tolerant effect. It converges to the steady-state error bound in a shorter time, thereby improving the stability and robustness of the humanoid robot.

[0034] 2. The finite-time path tracking control method of the humanoid robot has the ability to online adjust the weights and control parameters of the humanoid robot, has the characteristics of a simple motion model and weak coupling of degrees of freedom, and is not easily disturbed by the external environment during operation, thereby increasing the adaptive ability of the humanoid robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of a finite time path tracking control method for a humanoid robot proposed by the present invention;

[0036] Figure 2 This is a comparison diagram of the effects of the method of the present invention and the radial basis neural network method. DETAILED DESCRIPTION

[0037] 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.

[0038] The present invention provides a finite-time path tracking control method for a humanoid robot, which realizes a finite-time path tracking control of a humanoid robot by using an auto-disturbance rejection observer and a second-order sliding mode, comprising the following steps:

[0039] S1. Considering the contact force between the humanoid robot and the environment when working, the Newton-Euler equation is used to analyze the impedance control and control torque, and a matrix expression model of the nonlinear dynamics of the humanoid robot is established. The established matrix expression model is:

[0040]

[0041] Where: J is the transformation matrix between the inertial coordinate system and the mechanism coordinate system, M(θ e ) is a real symmetric inertia matrix, is the Coriolis force and centrifugal force matrix, H(θ e ) is the gravity matrix, η is the attitude angular displacement, and g(η) is the restoring force matrix;

[0042] S2, dividing the humanoid robot kinematic speed controller into an angular velocity subsystem controller and a position subsystem controller;

[0043] S3. According to the matrix expression model of the nonlinear dynamics of the humanoid robot, the angular velocity subsystem control model is designed so that the actual angular velocity can converge to the expected angular velocity within a limited time; the angular velocity subsystem control model is:

[0044] Where: c is the actual angular velocity, is the expected angular velocity, α is a positive number, k is a constant, θ e is the angle; for the matrix expression model of the nonlinear dynamics of the humanoid robot, the two variables α and k are adjusted, and the angular velocity subsystem control model is designed so that the actual angular velocity can converge to the desired angular velocity within a limited time.

[0045] For the angular velocity subsystem control model, select the integral sliding surface;

[0046] The selected integral sliding surface is:

[0047]

[0048] Where ξ is an auxiliary variable, λ1 and λ2 are both positive numbers, and tan(·) is the tangent function.

[0049] S4. Based on the actual angular velocity obtained from the matrix expression model of the nonlinear dynamics of the humanoid robot, the displacement subsystem model of the humanoid robot is designed:

[0050]

[0051] Where: v c is the actual linear velocity, is the expected linear velocity, x e and e represent the displacement of the humanoid robot in the x and y directions respectively;

[0052] S5. Design an auto-disturbance rejection observer for the displacement subsystem model of the humanoid robot:

[0053]

[0054] Where: tanh(·) is a kind of hyperbolic function, sig(·) is the Gaussian radial basis function, λ is the constraint force vector, N and I are the mass and inertia matrix of the humanoid robot, δ is the external disturbance, is the observation error; when the ADRO accurately estimates the external disturbance, the observation error In a neighborhood that converges to the origin in a finite time, the humanoid robot can estimate the external environmental interference in a finite time, which can effectively compensate for the system uncertainty caused by parameter perturbations and improve the stability and robustness of the humanoid robot.

[0055] S6. Design a control input vector τ for second-order sliding mode finite-time trajectory tracking so that the actual displacement of the humanoid robot can accurately track the given displacement;

[0056]

[0057] Where: σ and ρ are the driving torques generated by the servo motors of the humanoid robot, Γ(·) is the gamma function, r1 and r2 are positive constants;

[0058] S7, adjusting the control parameters and the control input vector of the second-order sliding mode finite-time trajectory tracking so that the tracking error of the humanoid robot completes minimum convergence within a finite time.

[0059] Choose the Lyapunov function as:

[0060] Where: Derivative V(τ) is used to adjust the control parameters. When By using the control input vector τ of the second-order sliding mode finite time trajectory tracking, the tracking error of the humanoid robot can be obtained to achieve minimum convergence in a finite time. Compared with the adaptive terminal sliding mode technology, this scheme has the ability to adjust the weights and control parameters of the humanoid robot online when used, has the characteristics of simple motion model and weak coupling of degrees of freedom, and is not easily disturbed by the external environment during operation, which increases the adaptive ability of the humanoid robot.

[0061] Example

[0062] The effect comparison diagram of the method of the present invention and the radial basis neural network method is as follows: Figure 2 As shown in the figure, it can be seen that compared with the radial basis neural network method, a humanoid robot finite time path tracking control method is implemented based on the self-disturbance rejection interference observer and the second-order sliding mode of the present invention, which can estimate the external environmental interference within a limited time, can effectively compensate for the system uncertainty caused by parameter perturbations, has a better fault tolerance effect, converges to the steady-state error bound in a shorter time, and improves the stability and robustness of the humanoid robot.

[0063] The standard parts used in the present invention can all be purchased from the market, and special-shaped parts can be customized according to the description in the specification and the drawings. The specific connection methods of each part adopt conventional means such as mature bolts, rivets, welding, etc. in the prior art. Machinery, parts and equipment all adopt conventional models in the prior art and will not be described in detail here. The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field. Although the embodiments of the present invention have been shown and described, it can be understood by ordinary technicians in this field that various changes, modifications, substitutions and modifications can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the attached claims and their equivalents.

Claims

1. A finite-time path tracking control method for a humanoid robot, characterized in that: A finite-time path tracking control of a humanoid robot is implemented by using an auto-disturbance rejection observer and a second-order sliding mode, including the following steps: S1. The impedance control and control torque are analyzed by Newton-Euler equation, and the matrix expression model of nonlinear dynamics of humanoid robot is established. The established matrix expression model is: Where: J is the transformation matrix between the inertial coordinate system and the mechanism coordinate system, M(θ e ) is a real symmetric inertia matrix, is the Coriolis force and centrifugal force matrix, H(θ e ) is the gravity matrix, η is the attitude angular displacement, and g(η) is the restoring force matrix; S2, dividing the humanoid robot kinematic speed controller into an angular velocity subsystem controller and a position subsystem controller; S3. Design an angular velocity subsystem control model for the matrix expression model of the nonlinear dynamics of the humanoid robot, so that the actual angular velocity can converge to the desired angular velocity within a limited time; select an integral sliding surface for the angular velocity subsystem control model; The angular velocity subsystem control model is: Where: c is the actual angular velocity, is the expected angular velocity, α is a positive number, k is a constant, θ e is the angle; S4. Based on the actual angular velocity obtained from the matrix expression model of the nonlinear dynamics of the humanoid robot, the displacement subsystem model of the humanoid robot is designed: Where: v c is the actual line speed, is the expected linear velocity, x e and e represent the displacement of the humanoid robot in the x and y directions respectively; S5. Design an auto-disturbance rejection observer for the displacement subsystem model of the humanoid robot: Where: tanh(·) is a kind of hyperbolic function, sig(·) is the Gaussian radial basis function, λ is the constraint force vector, N and I are the mass and inertia matrix of the humanoid robot, δ is the external disturbance, is the observation error; S6. Design a control input vector τ for second-order sliding mode finite-time trajectory tracking so that the actual displacement of the humanoid robot can accurately track the given displacement; Where: σ and ρ are the driving torques generated by the servo motors of the humanoid robot, Γ(·) is the gamma function, r1 and r2 are positive constants; S7, adjusting the control parameters and the control input vector of the second-order sliding mode finite-time trajectory tracking so that the tracking error of the humanoid robot completes minimum convergence within a finite time.

2. A finite-time path tracking control method for a humanoid robot according to claim 1, characterized in that: The Lyapunov function selected in S7 is: Where: Derivative V(τ) is used to adjust the control parameters. When By using the control input vector τ of the second-order sliding mode finite-time trajectory tracking, the tracking error of the humanoid robot can be minimized within a finite time.

3. A finite-time path tracking control method for a humanoid robot according to claim 1, characterized in that: According to the matrix expression model of nonlinear dynamics of humanoid robot, the two variables α and k are adjusted, and the angular velocity subsystem control model is designed so that the actual angular velocity can converge to the expected angular velocity within a finite time.

4. A finite-time path tracking control method for a humanoid robot according to claim 3, characterized in that: The selected integral sliding surface is: Where ξ is an auxiliary variable, λ1 and λ2 are both positive numbers, and tan(·) is the tangent function.

5. The finite-time path tracking control method of a humanoid robot according to claim 1, characterized in that: When the ADRO observer accurately estimates the external disturbance, the observation error Converges to a neighborhood near the origin in finite time.

6. The finite-time path tracking control method of a humanoid robot according to claim 1, characterized in that: The self-disturbance rejection observer and the second-order sliding mode are used to realize the finite-time path tracking control of the humanoid robot, which compensates the system uncertainty caused by parameter perturbations. It has the ability to adjust the weights and control parameters of the humanoid robot online, and has the characteristics of simple motion model and weak coupling of degrees of freedom.

Images