Design of Adaptive Neural Network Backstepping Controller for Linear Motor Magnetic Levitation System

Abstract

In order to solve the problem of poor anti-interference ability and long adjustment time of the magnetic levitation system of the controllable excitation linear synchronous motor, a backstepping control strategy based on adaptive RBF neural network is proposed. Firstly, according to the specific structure of CELSM, the working principle is analyzed, and the mathematical model of the magnetic levitation system is established. Secondly, the ideal backstepping controller of the magnetic levitation system is designed. The nonlinear untestable part of the controller is approximated by the RBF neural network, the neural network weight adaptive adjustment law is designed. Thirdly, the stability of the control system is proved by constructing the appropriate Lyapunov function. Finally, the control system is simulated and analyzed by Simulink. The simulation results show that the adaptive neural network backstepping controller designed in this paper the can effectively improve the performance of the magnetic levitation system of the controllable excitation linear synchronous motor and realize the stable operation of the linear synchronous motor magnetic levitation platform.