基于PSO的RBF神经网络的变频调速系统的研究

摘要

采用粒子群和RBF神经网络算法相结合的方法构造PSO-RBF神经网络PID控制器，利用S函数编写了MATLAB的PSO-RBF神经网络PID的M文件，并在SIMULINK环境下建立了基于PSO-BF神经网络PID的变频调速系统。仿真结果充分表明了该控制器具有良好的鲁棒性、跟随性和稳定性，而且改善了原系统的动态特性，证明了该方法在变频调速系统中的应用价值。%This paper adopt particle swarm algorithm and RBF neural network to construct the PID controller of PSO-RBF neural network , the M-Files of PSO-RBF neural network PID based on MAT-LAB through S-Function, and the mode of PSO -RBF neural network PID variable frequency speed control system was established in SIMULINK platform. Simulation results show that the controller hold well robustness, follow and stability,and the dynamic characteristics of the original system was improved, the application value of this method in the variable frequency speed control system was proved.