用于电阻炉的单神经元PID控制器的设计及仿真

本文对传统电阻炉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。介绍了基于有监督hebb学习算法，并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，该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，其各方面的控制效果均优于常规PID控制器。用于电阻炉温度控制系统的单神经元控制器较之传统的PID控制器能取得更好的控制效果。

Design and Simulation of Single-neuron PID Controller Used for Resistance Furnace

This paper analyzes the shortcomings of traditional PID controller used for resistance furnace, and expounds the advantages of single-neuron PID control algorithm. The single- neuron controller is introduced, which is designed based on supervisory hebb learning algorithm with combination of practical control experience. The meanings and dereferencing principle of each parameters in the single-neuron controller are analyzed and the single-neuron controller is emulated by using Matlab software under the circumstance of giving a step-input signal with the simulation contents of the connection weights, the influence of k value＇s variation on the system and its choosing method, and the comparison of their anti-jamming capabilities and regulating performances between the single-neuron PID controller and the conventional PID controller as well. The simulation result proves that the single-neuron PID controller has good parameter self-setting capability; better capacity of resisting disturbance and little overshoot, its control effect is superior to the conventional PID controller in all directions. The single-neuron PID controller .used for resistance furnace temperature control system can achieve better control effect than the traditional PID controller.