基于电阻炉的单神经元PID控制器的设计及仿真  

Controller design and simulation for resistance furnace based on Single-Neuron PID control

作　　者：方红[1] 王剑云[1] 

机构地区：[1]成都大学电子信息工程学院,成都610106

出　　处：《自动化与仪器仪表》2011年第5期55-56,60,共3页Automation & Instrumentation

摘　　要：对传统电阻炉PID控制器的不足之处进行了分析,阐述了单神经元PID控制算法的优点。介绍了根据有监督hebb学习算法,再结合实际控制经验而设计的单神经元控制器。分析了单神经元控制器中各个参数的意义与取值原则,并用Matlab软件对单神经元控制器在阶跃输入信号的情况下进行了仿真。仿真内容有:连接权值、K值变化对系统的影响及选取方法;单神经元PID控制器与常规PID控制器的抗干扰能力和调节性能对比。仿真结果证明单神经元控制器具有很好的参数自整定能力,且抗干扰能力强,超调量小,控制效果在各个方面都要优于常规PID控制器,单神经元控制器用于电阻炉温度控制系统比传统的PID控制器能取得更好的控制效果。This paper analyzes the inadequacies of the traditional PID controller for resistance furnace, and then it explains the advantages of single-neuron PID control algorithm. A single- neuron controller is proposed combined with practical experience after s supervised hebb learning algorithm was introduced. It analyzes the meaning of each parameter in single-neuron controller and its value principle. Simulation with Matlab sottware illustrates the results for a given step input. The results contain connection weights, the impact of K value change on system and its selection method, comparison of anti-interference ability and conditioning performance with single neuron PID controller and conventional PID controller. Simulation results show that the single-neuron controller has a good parameter self-tuning capability and anti-interference ability and a small overshot. The control effects in all aspects are better than conventional PID controller. They also show that a single-neuron controller for a resistance furnace temperature control system can achieve better control effect than the traditional PID controller.

关 键 词：单神经元PID HEBB学习规则 自适应控制器 仿真 

分 类 号：TP273.21[自动化与计算机技术—检测技术与自动化装置]