基于参数实时最优整定的智能PID控制器研究  

Research of Intelligent PID Controller Based on Parameters Real-time Optimal Tuning

作　　者：邓木生[1] 

机构地区：[1]湖南铁道职业技术学院,湖南株洲412001

出　　处：《计算机测量与控制》2011年第7期1629-1632,共4页Computer Measurement &Control

基　　金：湖南省科技计划基金资助项目(2009GK3186)

摘　　要：在参数时变系统中,为了解决PID参数不易实时调整问题,提出了基于PID控制律的智能控制方法;其主要思想是以PID的控制律作为神经网络输入输出模型,以PID的3个参数作为神经网络权值,通过对PID的控制模型进行实时在线训练,获得PID的最佳参数,从而实现对参数时变系统的最优控制;研究结果表明,基于PID控制模型的神经网络优化方法在处理非线性和时变系统时具有很强的鲁棒性,因而是一种有效的智能控制方法。The intelligent control approach based on PID control law is proposed in this paper for solving the problem to be difficult to adjust PID parameters in parameters time-varying system.The main idea is that the optimum control of parameters time-varying system is achieved with PID control model as input-output model of neural network,and with the three PID parameters as the weights of neural network,obtaining optimal PID parameters through the real-time online training of PID control model.The results show that the neural network optimization method based on the PID control model has strong robustness in dealing with nonlinear and time-varying systems.Hence,the method presented is an effective approach of intelligent control.

关 键 词：神经网络 PID控制器 非线性时变系统 参数实时优化 智能控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TP181[自动化与计算机技术—控制科学与工程]