基于神经网络的恒温控制系统的研究与设计  

Research on constant temperature control system based on neural network

作　　者：李泉溪[1] 巩庆民[2] 常波[2] 

机构地区：[1]河南理工大学计算机科学与技术学院,河南焦作454003 [2]河南理工大学电气工程与自动化学院,河南焦作454003

出　　处：《计算机工程与设计》2008年第11期2967-2969,共3页Computer Engineering and Design

摘　　要：由于恒温控制的数学模型很难建立,针对这一难题提出了基于神经网络的PID控制器,侧重介绍了BP神经网络PID控制器算法的基本知识以及控制器的设计原理,通过实验仿真证明了神经网络PID控制器的控制效果比传统的PID控制在静态特性和动态特性方面都有所提高,而且具有较好的鲁棒性,该控制器在工业控制中将会发挥越来越大的作用。The model of constant temperature control system is difficult to establish. In order to slove this problem, a kind of PID controller based on neural network is presented. The basic knowledge of learning arithmetic about neural network and design principle of PID based on neural network are proposed. The experimental result show that the PID controller based on neural network provides an excellent control result than the traditional PID controller. The robustly stable is excellent, This kind of controller will be widely used in the industry.

关 键 词：神经网络 比例、积分、微分控制 BP网络 参数整定 恒温控制 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]