适用于电机辨识建模的改进差分进化算法  

An Improved Differential Evolution Algorithm for Motor Identification Modeling

作　　者：黄文文 宋璐[1] 史敬灼[1] HUANG Wen-wen;SONG Lu;SHI Jing-zhuo(Henan University of Science and Technology,Luoyang 471023,China)

机构地区：[1]河南科技大学

出　　处：《微特电机》2019年第11期34-37,共4页Small & Special Electrical Machines

基　　金：国家自然科学基金项目(U1304501)

摘　　要：针对采用标准差分进化算法进行超声波电动机辨识建模存在的算法不够稳健、建模效率低等问题,对差分进化算法的变异操作进行改进,并设计自适应系数来调节优化进程。电机模型辨识应用表明,该改进差分进化算法表现更稳健,建模精度和建模效率更高,更适合电机辨识建模应用。Aiming at the problems that the algorithm for the ultrasonic motor identification modeling using standard differential evolution algorithm is not robust enough and the modeling efficiency is low,the mutation operation of the standard differential evolution algorithm was improved,and the adaptive coefficient was designed to adjust the optimization process.The application of motor model identification shows that the improved differential evolution algorithm is more robust,with higher modeling accuracy and efficiency,and is more suitable for motor identification modeling applications.

关 键 词：超声波电动机 辨识建模 差分进化算法 

分 类 号：TM359.9[电气工程—电机]