摘要
在大型火力发电厂中,由于主汽温对象具有大延迟、大惯性、非线性以及时变性的特性,导致了对其控制比较困难。将PID控制器和具有学习功能的神经网络相结合,是智能控制的一种新的途径。本文针对超临界直流锅炉主汽温对象的调节特点,将中间点温度信号作为前馈引入主汽温调节系统,利用神经网络进行系统辨识,运用单纯形法寻优PID控制参数,达到最优控制效果。
In modem thermal power plant, because of the big delay of the main steam temperature object＇s characteristic, big inertial, nonlinear and the variety model with variety time, which causing as to it＇s control more difficult. The combination of the broad-using PID controller and ANN is a way of intelligent control. This paper inducts the intermediate point temperature to the main steam control system as feed forward signal based on its effect.Then optimizes PID parameters with intelligent control technologies and distingnish system with neural network and the simplex search method is used to seek excellent PID control parameters and get the optimum result of control.
出处
《微计算机信息》
北大核心
2007年第06S期34-36,共3页
Control & Automation
关键词
神经网络
系统辨识
直流炉
单纯形法
PID控制
neural network, system identification, once-through boiler, simplex search method, PID control
