Robust Adaptive Control Of Robotic Manipulators Based On Terminal Sliding Mode

High-accuracy tracking control is a research hotspot in the field of robotic manipulator control.In terms of the position tracking control problem of the manipulators whose end-effectors are not in contact with constrained surfaces,the reduced-order terminal sliding mode(ROTSM)has raised much attention,since it combines the superior features of both sliding mode control and finite-time control.The chattering and the singularity problems,however,often exist in conventional ROTSM controllers.The existing methods to solve these problems often cause the tracking errors to lose the finite-time convergence to zero.In terms of the simultaneous position and force tracking control problem of the manipulators whose end-effectors are constrained,most of the existing researches are focused on the linear sliding mode(LSM)control.In spite of the superior features of terminal sliding mode(TSM),the dynamical coupling between the motion and the constraint force hinders the application of TSM to the simultaneous position and force tracking control.Therefore,there is still much work to do in the field of TSM-based position/force control.This dissertation considers the problem of high-accuracy tracking of robotic manipulators.Based on a series of control theories,such as TSM control,position/force control,robust control,neural network control and Lyapunov stability theory,a set of reduced-order and full-order TSM-based tracking control strategies are proposed.The main work is organized as follows.(1)For the position tracking problem of the manipulators whose end-effectors are not in contact with constrained surfaces,a series of robust adaptive control strategies are proposed based on the full-order terminal sliding mode(FOTSM).Firstly,a kind of FOTSM manifold is built based on the tracking errors.Then a robust algorithm and two radial basis function(RBF)neural network algorithms are designed to overcome the system uncertainties.The proposed methods can drive the tracking errors to reach the FOTSM and then converge to zero in finite time.Compared with the existing ROTSM control methods,the proposed methods can avoid the chattering and the singularity problems while ensuring the finite-time convergence of the tracking errors.Simulation results verify the effectiveness and the advantages of the proposed methods in comparison with the existing position tracking control methods.(2)For the simultaneous position and force tracking problem of the manipulators whose end-effectors are in contact with constrained surfaces,a series of ROTSM-based robust adaptive control strategies are proposed.In terms of the position tracking,a ROTSM manifold is built based on the position and the velocity tracking errors.Then a robust algorithm and a RBF neural network algorithm are designed to guarantee the finite-time convergence of the position and velocity tracking errors.In terms of the force tracking,a force control gain is introduced to guarantee the boundedness of the force tracking error and adjust the boundary.Compared with the existing LSM-based position/force control methods,the proposed methods can guarantee the finite-time convergence of the position and velocity tracking errors and enhance the steady-state tracking precision.Simulation results verify the effectiveness and the advantages of the proposed methods in comparison with the existing position/force tracking control methods.(3)For the simultaneous position and force tracking problem of the manipulators whose end-effectors are in contact with constrained surfaces,a framework of ROTSM is built and a series of robust adaptive control strategies are proposed.The framework introduces the ROTSM control concept into both the position tracking and the force tracking.A new expansion of ROTSM is built and its finite-time convergence is analysed.Compared with the existing position/force control methods,the proposed methods can guarantee the finite-time convergence of both the position and the force tracking errors and enhance the steady-state tracking precision.Simulation results verify the effectiveness and the advantages of the proposed methods in comparison with the existing position/force tracking control methods.(4)For the simultaneous position and force tracking problem of the manipulators whose end-effectors are in contact with constrained surfaces,a series of robust adaptive control strategies are proposed based on the FOTSM.Firstly,a dynamic model of the constrained manipulator including actuator dynamics is built and a new expansion of FOTSM is designed.The system uncertainties are overcome by the robust control and the neural network control.Compared with the ROTSM-based position/force control methods,the proposed methods can avoid the chattering and the singularity problems while ensuring the finite-time convergence of the tracking errors.Simulation results verify the effectiveness and the advantages of the proposed methods in comparison with the existing position/force tracking control methods.Through the above research,a set of TSM-based robust adaptive tracking control strategies are designed for the robotic manipulators whose end-effectors are constrained or not constrained.The proposed strategies enhance the steady-state tracking precision and the robustness with respect to system uncertainties,and provide a new way to the research on the high-accuracy tracking control of robotic manipulators.