Research On Trajectory Tracking Mode-based And Mode Free Control Methods Of Manipulator System

With the rapid development of intelligent manufacturing in China,the research and development level of robotic manipulator has grown rapidly,making the precise tracking control of robotic arm gradually become a research hot issue.In this paper,the trajectory tracking control of multi-input and multi-output n-degree-of-freedom manipulator system is studied based on model and model-free control.At the same time,combining with the model uncertainty,disturbance,and other nonlinear problems of the manipulator,such as output constraints and event triggering,the controller is designed.The stability of the control system is proved theoretically by means of Lyapunov equation and the effectiveness of the proposed method is verified by simulation.Firstly,the dynamic modeling and basic control methods of manipulator are introduced.The Euler-Lagrange equation of the manipulator is derived by taking the two-link manipulator system as an example,and is generalized to the dynamic equation of the manipulator system of n degrees of freedom.At the same time,this paper introduces the design ideas of four control methods: adaptive control,neural network control,iterative control and model-free adaptive control,which lays a theoretical foundation for the design of the later controller.Secondly,we study event-triggered based adaptive neural network(NN)tracking control of a robotic manipulator with output constraints and disturbance.A novel asymmetric tan-type barrier Lyapunov function(BLF)is developed to satisfy the requirement of time-varying output constraints.Then,a fixed threshold event triggering is proposed to reduce the energy consumption,which avoids the happening of Zeno behavior after analysis.Further,a disturbance observer(DO)and an adaptive neural network are devised to estimate the bounded disturbance and the unknown dynamics of the robotic manipulator.The proposed controller can achieve uniform boundness of the solution and adjustment of transient performance.The event triggering adaptive neural network control,terminal sliding mode control and event triggering times are simulated and discussed by using 2 degrees of freedom manipulator systemFinally,we devote to the tracking control problem for a robotic manipulator with the model free adaptive iterative control(MFAIC)approach.The dynamical model of manipulator is converted into a linear discrete model by compact form dynamic linearization(CFDL)with iteration method.Then,based on the linear model,a MFAIC algorithm is proposed to control the system tracking the desired time-varying trajectories,which just depends on the iterative input and output data of each joint of the manipulator.It is proved that the convergence of both tracking error and time-varying parameters can be guaranteed with the increasing of time and iterative numbers.The effectiveness and superiority of the presented method is verified by related simulation results.