三容系统的解耦神经网络PID控制  

Neural Network PID Control of Three-Tank System Based on Decoupling

作　　者：葛锁良[1] 卢娟[1] 

机构地区：[1]合肥工业大学电气与自动化工程学院,安徽合肥230009

出　　处：《湖南工业大学学报》2008年第5期14-17,共4页Journal of Hunan University of Technology

摘　　要：三容水箱是常见的非线性强耦合系统,利用一种非线性动态解耦的方法对其进行解耦(其特点是能利用非线性补偿的方式将该类系统各回路输入与输出之间完全解耦),然后采用基于BP神经网络的PID控制策略对其进行控制,并利用MATLAB进行仿真。仿真结果证明解耦的有效性,表明了BP神经网络的PID算法比传统的数字PID具有更强的抗扰动性。A decoupling method for non-linear dynamic system is applied in the familiar non-linear coupled three-tank water system.（Its advantage is that absolute decoupling for the nominal responses of system outputs can be implemented for this kind of system which takes a non-linear compensation way.） The PID control based on BPNN is taken in to this system and simulated in MATLAB. The result proves the efficiency of decoupling method, and shows this system has stronger antidisturbance characteristic under the control of PID based on BPNN than that of Digital PID.

关 键 词：三容水箱 神经网络 PID 仿真 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]