参数自整定2自由度PID全神经元实现的仿真研究  

Full Neuron Realization of Parameters Auto-Adjusting Two Degree-of-Freedom PID

作　　者：邱公伟[1] 林瑞全[1] 

机构地区：[1]福州大学电气工程系,福建福州350002

出　　处：《系统仿真学报》2002年第10期1293-1295,1309,共4页Journal of System Simulation

基　　金：福建省自然科学基金项目(E0010017)资助

摘　　要：2自由度PID控制器参数整定是一个复杂而又困难的问题。本文采用全神经元方法实现了2自由度PID控制,通过权值自学习解决了参数自动整定这一难题。只要选择适当的神经元权值系数,就可以是系统的抗干扰能力和跟踪给定的能力同时达到最佳,并使系统具有自适应能力和较强的鲁棒性。Adjusting parameters of Two Degree-of-Freedom PID controller is a very difficult problem.In this paper, we propose a full neuron realization method of two degree-of-freedom PID controller,and the difficult problem for auto-adjusting parameters is solved by weight value auto-learning.By selecting proper coefficient of neuron weight, the ability of rejecting disturbance and tracking set value can get better at the same time, also the system has self-adaptive ability and higher robustness.

关 键 词：参数自整定 2自由度PID控制 神经元 仿真 神经元控制器 工业控制系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TP391.9[自动化与计算机技术—控制科学与工程]