Research On Trajectory Tracking Control Method Of Manipulator Based On Sliding Mode Variable Structure

Manipulator is a complex MIMO controlled system with strong coupling,high nonlinearity,model uncertainties and external interference.The realization of fast and accurate trajectory tracking has always been a difficult point in the field of manipulator,and it is also a research hotspot.In this paper,based on the study of kinematics and dynamics of the series manipulator,considering the uncertainty of the manipulator model and the existence of external interference,the trajectory tracking control strategy of the manipulator based on sliding mode variable structure is studied,the main tasks are as follows:1.Taking the RBT-6T/S03 S manipulator as the research object,the kinematics model of the manipulator is established by using the standard D-H parameter method.The forward kinematics equation is derived,and the inverse kinematics solution is obtained by analytical method.The Lagrangian modeling method is used to establish a simplified RBT-6T/S03 S manipulator dynamic model,which provides a test object for the simulation experiment of trajectory tracking control strategy.2.The trajectory planning method of manipulator is studied.The third-degree polynomial and fifth-degree polynomial trajectory planning algorithms in the joint space are deduced in detail.Simulation comparison analysis shows that the fifth-degree polynomial planning has better performance in polynomial interpolation trajectory planning with intermediate path points.The linear interpolation algorithm and circular interpolation algorithm in Cartesian space are derived.The rationality and effectiveness of the interpolation algorithm are verified by simulation experiments.3.Considering the modeling error and friction of the manipulator system,the RBF neural network is used to approximate the model uncertainty consisting of the modeling error and friction.Designed a sliding mode controller based on exponential reaching law,and a robust term is added to the sliding mode controller to offset the approximation error of the RBF neural network.Based on the Lyapunov stability criterion,the weight update law of the neural network is derived,the online adjustment of network weights can realize the adaptive approximation of neural network.Theoretical analysis and simulation comparison experiment verify the rationality and superiority of the control strategy.4.Considering the modeling error,friction and external disturbance of the manipulator system,a nonlinear disturbance observer that does not require the feedback signal of joint angular acceleration is designed.The uncertain term of the manipulator model and the external disturbance are combined into a composite disturbance,and the nonlinear disturbance observer is used to estimate the composite disturbance and compensate it at the control input.An improved power reaching law is selected to improve the dynamic quality of the approaching motion,and a robust term that offsets the estimation error of the nonlinear disturbance observer is added to the nonsingular fast terminal sliding mode controller to enhance the robustness of the system.The Lyapunov function is used to prove the stability of the system.Theoretical analysis and simulation comparison experiment verify the rationality and superiority of the control strategy.