纯滞后对象的RBF神经网络PID控制研究

PID Controller Based on RBF Network

神经元网络具有信息分布储存、并行处理以及自组织、自学习能力等优点。运用RBF神经网络训练了PID控制器的三个重要参数。相比于传统的PID控制器,提高了控制精度,使系统快速地达到了稳定状态。并进行了Matlab仿真,验证了这一结论。

Neural network has many advantages such as distributing memory of information,parallel processing and ability of self-organizing and self-studying.This paper trains the three important parameters of PID controller by using RBF neural network.Compared with traditional PID controller,it improves the control precision and makes the system reach to steady state rapidly.And it proves this conclusion by Matlab simulation.