Search

CN-121989230-A - Flexible joint mechanical arm tracking control method based on UDE

CN121989230ACN 121989230 ACN121989230 ACN 121989230ACN-121989230-A

Abstract

The invention provides a UDE-based flexible joint mechanical arm tracking control method, which is applied to the field of robot intelligence and comprises the steps of defining a gap hysteresis nonlinear phenomenon based on a gap hysteresis phenomenon of a robot dynamics model; based on the time derivative of the tracking error and the expected error dynamics characteristic of the filtering error, combining a robot dynamics model and a similar clearance lag nonlinear phenomenon to obtain a dynamics equation carrying an uncertainty term, approximating the uncertainty term to obtain an estimated value of the uncertainty term, and combining the dynamics equation carrying the uncertainty term based on the estimated value of the uncertainty term to obtain a control model of the controller based on uncertainty and disturbance estimation. The invention refers to modeling errors, parameter uncertainty and hysteresis nonlinear characteristics as total disturbance, thereby avoiding the complexity of modeling each interference source one by one commonly seen in the traditional method, compensating disturbance in real time by simply adjusting bandwidth parameters of the UDE filter, and remarkably improving control precision.

Inventors

  • GAO HEJIA
  • ZHAO YUANYUAN
  • XIAO XU
  • ZHOU JUNZE

Assignees

  • 安徽大学

Dates

Publication Date
20260508
Application Date
20251231

Claims (9)

  1. 1. A UDE-based flexible joint mechanical arm tracking control method is characterized by comprising the following steps: S1, constructing a robot dynamics model based on a Dener-Hartab parameter method; s2, defining a gap hysteresis non-linear phenomenon based on a robot dynamics model; S3, based on the time derivative of the tracking error and the expected error dynamics characteristic of the filtering error, combining a robot dynamics model and a similar gap hysteresis nonlinear phenomenon to obtain a dynamics equation carrying an uncertainty term; s4, approximating the uncertain item to obtain an estimated value of the uncertain item; and S5, combining the dynamic equation carrying the uncertainty term based on the estimated value of the uncertainty term to obtain a control model of the controller based on uncertainty and disturbance estimation.
  2. 2. The UDE-based flexible joint manipulator tracking control method according to claim 1, characterized in that the robot dynamics model in S1 is formula (1): (1) Wherein, the A generalized coordinate vector for the robot link position; Time derivative of ρ Is a velocity vector; Time derivative of ρ Is an acceleration vector; The inertia matrix is positively determined for symmetry; Is the coupling effect of coriolis force and centrifugal force; a gravity potential gradient vector; The input vector is lagged for the gap.
  3. 3. The UDE-based flexible joint manipulator tracking control method according to claim 2, characterized in that the expression of the gap hysteresis of the robot dynamics model in S2 is formula (2): (2) nonlinear characteristics of the input for the ith gap hysteresis; t is time; is the i first constant; is the i-th input vector; Is the slope of the ith line; is the ith second constant; , for the set of slopes of all the straight lines, Is a set of all second constants.
  4. 4. The UDE-based flexible joint manipulator tracking control method according to claim 3, wherein the defining a gap-like hysteresis nonlinear phenomenon in S2 includes: s202, solving the expression of the clearance hysteresis of the robot dynamics model Segmentation monotonic explicit analytical solution, with respect to The piecewise monotonic explicit analytical solution includes equation (3) and equation (4): ;(3) is the i-th input vector; is a process parameter; (4) An initial value for the i-th gap lag input; an initial value for the i-th input vector; A derivative of the ith input vector; is an input vector; S203, solving the clearance hysteresis phenomenon based on the robot dynamics model and related to The piecewise monotonic explicit analytical solution is used for obtaining an expression for defining the class gap hysteresis nonlinear phenomenon, wherein the expression for defining the class gap hysteresis nonlinear phenomenon is expressed as a formula (5): (5) Wherein, the , ; Is a process parameter and , ; Wherein B is a constant matrix; is a process parameter; Is that Is a maximum value of (a).
  5. 5. The UDE-based flexible joint manipulator tracking control method according to claim 4, wherein the obtaining a dynamic equation carrying an uncertainty term based on the time derivative of the tracking error and the expected error dynamic characteristic of the filtering error in S3, in combination with a robot dynamic model and a gap-like lag nonlinear phenomenon, includes: s301, defining a tracking error and a time derivative of the tracking error; S302, defining a filtering error and expected error dynamics characteristics of the filtering error; S303, obtaining an simultaneous robot dynamics model, an expression of a similar gap hysteresis nonlinear phenomenon, a time derivative of a tracking error and an expected error dynamics characteristic of a filtering error, and obtaining a formula (6): (6) Is the coupling effect of coriolis force and centrifugal force; a gravity potential gradient vector; A positive gain matrix for error feedback; Is a constant matrix; time derivative of tracking error; is a tracking error; Is a desired trajectory; Is the derivative of the desired trajectory; s304, obtaining a dynamic equation carrying an uncertain term based on the formula (6), wherein the dynamic equation carrying the uncertain term is shown as the formula (7): (7) g is a control parameter; in order to determine the term(s) of uncertainty, 。
  6. 6. The UDE-based flexible joint manipulator tracking control method according to claim 5, wherein approximating the uncertainty term in S4, obtaining an estimate of the uncertainty term includes: setting a filter At the position of The unity gain and zero phase offset are maintained in the main band and zero gain is maintained in the other bands, resulting in an estimate of the uncertainty term.
  7. 7. The UDE-based flexible joint robot arm tracking control method according to claim 6, characterized in that the expression of the estimate of the uncertainty term is formula (8): (8) an estimate of an uncertainty item; inverse laplace transform; is a convolution operation.
  8. 8. The UDE-based flexible joint manipulator tracking control method according to claim 7, wherein the estimating based on the uncertainty term in S5 combines a kinetic equation carrying the uncertainty term, and obtaining a control model of the controller based on the uncertainty and disturbance estimation includes: S501, combining the formula (7) based on the formula (8), and deriving the formula (9): (9) S502, performing phenotype on the formula (9) to obtain a control model of the controller based on uncertainty and disturbance estimation, wherein the control model of the controller based on uncertainty and disturbance estimation is shown as a formula (10): (10)。
  9. 9. the UDE-based flexible joint manipulator tracking control method according to claim 7, wherein in S502, when the velocity measurement is not available, a complete state vector cannot be directly obtained The control model of the controller is formula (11): (11)。

Description

Flexible joint mechanical arm tracking control method based on UDE Technical Field The invention relates to the field of robot intelligence, in particular to a tracking control method of a flexible joint mechanical arm based on UDE. Background In the field of robot research, track tracking is always a core topic in mechanical arm control, and related technologies thereof are widely applied to various scenes such as aerospace, automobile manufacturing, industrial automation and service industries. The conventional rigid mechanical arm is stable in performance when performing high-repeatability and high-precision tasks, but generally can only strictly follow a preset track due to limited freedom, so that the adaptability in complex or dynamic environments is insufficient. In addition, such mechanical arms often lack compliance when interacting with humans or other dynamic objects, easily cause potential safety hazards, and affect overall operating efficiency. With the rapid development of automation technology and artificial intelligence, the requirement on the track tracking capability of the mechanical arm is increasingly increased. Application scenes have also gradually expanded from highly structured industrial environments to unstructured scenes that are closer to the real world. To cope with this trend, flexible articulated robotic arms are increasingly gaining attention. Compared with the traditional rigid structure, the mechanical arm has more degrees of freedom, can simulate the actions of hands more realistically, and has better performance in the aspects of flexibility and environmental adaptability. However, the dynamic model of the flexible joint robot is complex, and the nonlinearity and strong coupling characteristics of the flexible joint robot also bring challenges to the design of the controller, so that the development of an efficient control algorithm matched with the flexible joint robot is particularly critical. Under the scene of smaller fluctuation of ideal working conditions and system parameters, the traditional control method can still meet the basic performance requirement. However, the method has a plurality of inherent limitations that firstly, the effectiveness of the method is seriously dependent on the accuracy of a system model, and for complex objects with high nonlinearity and strong coupling, the accurate modeling is often difficult to realize, and key dynamic characteristics are easy to lose by simplifying the model, secondly, the traditional theory is mainly based on the assumption of a linear steady system, and the nonlinear, time-varying characteristics and uncertainty commonly existing in an actual system are difficult to effectively process, thirdly, the traditional method often faces challenges of insufficient instantaneity and limited adaptability in complex control tasks such as multivariable coupling, constraint processing, rapid time-varying dynamics and the like. Especially when there is significant nonlinearity, model uncertainty, or severe change in operating point of the system, its control performance will drop significantly. In addition, the traditional control system is generally complicated in parameter setting process and poor in expansibility. As such, in the face of increasingly complex dynamic systems of modern day, conventional approaches often need to be combined with advanced control strategies to improve overall control performance and environmental flexibility. Disclosure of Invention In order to solve the problems, the invention provides a tracking control method of a flexible joint mechanical arm based on UDE, which explicitly considers the recoil hysteresis effect in a system modeling and control frame and provides an effective solution for the nonlinear problem caused by an elastic component in the flexible joint mechanical arm, wherein the method comprises the following steps: A method for controlling tracking of a flexible joint mechanical arm based on UDE, the method comprising: S1, constructing a robot dynamics model based on a Dener-Hartab parameter method; s2, defining a gap hysteresis non-linear phenomenon based on a robot dynamics model; S3, based on the time derivative of the tracking error and the expected error dynamics characteristic of the filtering error, combining a robot dynamics model and a similar gap hysteresis nonlinear phenomenon to obtain a dynamics equation carrying an uncertainty term; s4, approximating the uncertain item to obtain an estimated value of the uncertain item; and S5, combining the dynamic equation carrying the uncertainty term based on the estimated value of the uncertainty term to obtain a control model of the controller based on uncertainty and disturbance estimation. Optionally, the robot dynamics model in S1 is formula (1): (1) Wherein, the A generalized coordinate vector for the robot link position; Time derivative of ρ Is a velocity vector; Time derivative of ρ Is an acceleration vector; The inert