自适应神经网络PID控制器

Neural network PID controller of the self-adaptation

提出了一种新型神经网络PID控制器,其学习速率是通过三层前向BP网络在线辨识学习,使神经元有较强的智能性、自适应和自学习的能力;同时,将Smith预估器与神经元PID控制器相结合,能更有效地抑制纯滞后的影响。仿真结果表明该控制器有较好的控制效果和鲁棒性。

This text put forward a kind of new-type neural network PID controller,the speed of study of this controller distinguished and studied online to BP network before three layers,Make neuron have a strong one intelligent,self-adaptation and from the ability of study;The estimating Smith in advance device combine with neuron PID controller,can suppress the pure influence that lags behind more effectively.The emulation result shows:this controller has better control result and stupid and excellent.