Design And Application Of Variable Impedance Control For Uncertain Robot Systems

Since the appearance of robot system,there has been considerable development and a lot of research.Robot system is composed of robot,operating object and environment.External conditions,measurement errors or other reasons will make it difficult to establish an accurate dynamic model.The basic starting point of this thesis is the control of uncertain robot system.As for robot control,one of the basic requirements is to improve the interaction performance between robot and the environment,so that it can interact with various environments in a stable way at the same time.Therefore,in order to improve the human-computer interaction performance of uncertain robot system in different environments,the research of variable impedance control problem is the focus of this thesis.This thesis discusses the design and application of variable impedance control for an uncertain robot system.On this basis,an adaptive neural network control scheme is designed to estimate the unknown model in robot system.The stability of the whole closedloop system was proved by Lyapunov direct method,and the design method was applied to other manipulator systems.Finally,the effectiveness and feasibility of the system were verified by simulation software.Firstly,a variable impedance control scheme is proposed.This control method is helpful for the robot to complete the given interactive task in the unknown environment,and improves the overall performance of the robot-environment system.Stiffness,damping,and inertia can be varied during the interaction task.When the inertia,stiffness and damping matrices are properly selected,the tracking error can approach zero,and the effectiveness of this scheme can be proved by simulation results.Secondly,the adaptive neural network control scheme is designed.The Lyapunov function is designed,and the ideal controller of the system is derived by iterative Backstepping design.The unknown part is approximated by the constructed adaptive neural network,and then the actual usable control signals are obtained.Furthermore,the adaptive parameters are constructed.The uniform boundedness of the closed-loop system is proved by Lyapunov direct method,and the tracking error tends to zero.The tracking performance of this scheme is illustrated by simulation results.Finally,the proposed variable impedance control scheme and the adaptive neural network control design scheme based on the proposed scheme are applied to other manipulator systems to transform the control problem of the manipulator system into a tracking control problem.Under the premise of the designed control scheme,it can perfectly track the expected reference trajectory.The interaction performance between the robot arm system and the environment can be improved by the designed variable impedance control method,the adaptive neural network control method can estimate the unknown model in the robot system,and the uniformly bounded performance of the robot arm system can be guaranteed.In summary,the variable impedance control strategy is designed for the uncertain robot system.The unknown robot dynamics model in the system is approximated by an adaptive neural network control scheme designed based on this strategy.In order to verify whether the scheme designed in this thesis is feasible,the scheme is applied to other robotic arms.