RBF NEURAL NETWORK PID SPACE VECTOR CONTROL OF LINEAR SERVO LOAD SIMULATOR

摘要

Aiming at the complex structure and low reliability of the traditional linear servo load simulator that converts the rotary motion into linear motion, a simple and stable direct loading line of the permanent magnet synchronous linear motor is designed. Due to the traditional proportion integration differentiation (PID) controller, control parameters cannot be adjusted by the environment; so this paper improves the traditional PID controller, an online self-tuning PID vector control method based on radial basis function (RBF) neural network parameter optimization is proposed, modelled and simulated by MATLAB/Simulink. The motor and its space vector control system were modelled and simulated. The simulation results show that the PID control based on RBF neural network has optimal dynamic response and more stable tracking performance. The experimental results also prove the feasibility and effectiveness of the proposed method.