基于单神经元PID控制器在渠道自动控制中的应用  

Application of neural network PID controller for canal control system

作　　者：王涛[1] 阮新建[2] 

机构地区：[1]深圳市南山区农业水务局,广东深圳518000 [2]武汉大学水资源与水电工程科学国家重点实验室,湖北武汉430072

出　　处：《长江科学院院报》2004年第4期53-56,共4页Journal of Changjiang River Scientific Research Institute

基　　金：国家863计划项目(2001AA242111)

摘　　要：将基于单神经元自适应PID控制器引入到渠道系统自动控制中,它实现了PID控制中参数整定不依赖于渠道数学模型,且能在线依据渠道系统的动态信息调整PID参数,以满足适时控制的要求。通过Matlab软件的仿真,结果表明:采用基于单神经元自适应PID控制器,渠道运行响应加快,水位超调变小,适应性增强,且有更好的动态和稳态性能。Applying adaptive neural network PID Control theory, a controller is designed for channel operation system.The concordance of the parameters in the PID control does not depend on the channel mathematical model, and it can regularize PID parameters on line from dynamic information of channel system to meet the requirements of in time control. Through the simulation of Matlab software, the result shows that the adaptive PID controller possesses the advantages of quick response in operation, small over regulation magnitude, good self suitabitity, and good dynamic and stability performances.

关 键 词：渠道 自动化 PID控制 神经网络 数学模型 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]