用遗传算法建立连续系统低阶近似模型的方法及其在电机控制中的应用

A METHOD TO BUILD LOW-ORDER APPROXIMATE MODEL OF CONTINUOUS SYSTEMS BY USING GENETIC ALGORITHM AND ITS APPLICATION TO MOTOR VARIABLE SPEED SYSTEMS

基于高阶非线性系统的阶跃响应数据，用遗传算法寻找最优参数，直接得到便于工程设计和分析的低阶线性连续系统模型．首先讨论了遗传算法及其改进方法，然后分析低阶系统参数变化范围的自动修改规则和建模方法，最后建立了交流电机调速系统的低阶近似模型．

ow--order linear systems are convenient for engineering design and analysis. This paper presents a method to build low--order linear models of high-order nonlinear systems.The method is based on the step response data of high-order nonlinear system and uses genetic algorithm to search the optimal parameters. First, the genetic algorithm and thewaysto improve it are discussed; Then, the modeling methods and the rules to alter theparameter ranges are presented; At last, a low-order approximate model of AC motorvariable speed systems is obtained by using the presented method.