Control Of A Multi-joint Robot Driven By A Permanent Magnet Synchronous Motor With Unknown Load

The safety,stability and fast tracking of multi-joint articulated robot system driven by permanent magnet synchronous motors have been the focus of nonlinear control research.Therefore,it has important research value to realize fast and smooth tracking of the system when there are modeling errors,external loads and constraints.The research content of this paper is as follows.(1)To solve the problems of modeling error,external disturbance and unknown load,a dynamic surface control strategy of radial basis function neural network with Sage-Husa adaptive Kalman filter load torque observer is designed.Firstly,the load torque observer is designed by using the Sage-Huga adaptive Kalman filter.Secondly,the radial basis function neural network dynamic surface strategy is designed to deal with the modeling errors and external disturbance of systems,and “computational complexity” of the backstepping control when applied to high order system.The stability of this system is demonstrated by the performance analysis of the control system.Finally,comparing the proposed control strategy with the common control strategies,the simulations and experiments show that the control strategy can track the desired trajectory within 0.21 s and reduce the steady-state error range to ±0.00086 rad.(2)To solve the problems of modeling error,external load and output constraints in the system,the smooth-switching for backstepping gain control strategy based on the asymmetric time-varying Barrier Lyapunov Function and adaptive neural network is proposed.Firstly,the adaptive neural network is designed to approximate modeling errors,unknown loads and unenforced inputs in the system.Secondly,the gain function based on the error and the rate of error change are designed to improve the response speed and tracking stability.Combining the advantages of the two time-varying gain functions,the smooth-switching backstepping gain strategy based on Barrier Lyapunov function is designed.The stability of the proposed control system,output constraint and tracking error convergence of the system are verified by the system performance analysis.The simulation results show that the output of the proposed strategy can be kept within the output constraints,and the rise time and steady-state error range of the system are effectively reduced to 0.1894 s and ±0.0005 rad.(3)This strategy focuses on the design of smooth-switching gain dynamic surface control based on adaptive fuzzy neural network for articulated robot driven by permanent magnet synchronous motors with input voltage saturation and asymmetric time-varying full-state constraints.The auxiliary system is designed to handle the problems of the input saturation function is not smooth.The asymmetric time-varying Barrier Lyapunov Function is established to handle the problems of the full-state constraints.The adaptive fuzzy neural network is designed to approximate the modeling error of the system.Additionally,two gain functions are designed to improve the rapidity and smoothness of tracking,the dynamic surface control is designed with the smooth-switching gain strategy.The stability of the control system,the constraint of all states and the convergence of all signals are verified by the performance analysis of the system.Finally,the simulation results show that the maximum tracking error of the joint of the articulated robot is ±0.00005 rad,and the proposed strategy effectively reduces the tracking error of the manipulator system by0.00002 rad.The three strategies proposed above use load torque observer,adaptive network and adaptive fuzzy network to study the compensation strategies of the system modeling error and external load,and solve the influence of modeling error and external load on the position tracking control of the articulated robot system.The asymmetrical time-varying barrier Lyapunov function is used to study the output and state constraint strategy of articulated robot system,and the asymmetrical time-varying constraints of system output and state are solved.The smooth-switching gain strategy is used to study the influence of control gain on tracking performance,which solves the problem that it is difficult to improve the rapidity and smoothness of tracking control simultaneously.This research has great significance in tracking control of the multi-joint articulated robot system driven by permanent magnet synchronous motors.