纯滞后系统的神经元PID和Smith预估器控制

Single Neuron PID Control with Smith Predictor in Pure Lag System

PID 控制器具有原理简单,鲁棒性好,稳定性高而得到广泛应用。但对于纯滞后系统,PID 控制很难获得理想的控制效果,甚至产生不稳定。为此,采用了单神经元 PID 控制器,利用其自学习,自组织能力,根据被控对象的变化情况对控制器的权值进行在线调整,构成了自适应控制,结合 Smith 预估器对纯滞后系统进行控制,取得良好效果。

PID controller has the features of simple principle,good robustness and high stability,and so it has been widely used.However,it is greatly difficult for PID control to acquire ideal control effect to pure lag system,even causes unstable of system.By use of self-study and self-organization ability and according to the change of controlled process,the weights of single neuron PID controller are on-line regu- lated,making up adaptive control from this.This controller being combined with smith predictor controls the pure lag system,bring about good results.