中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊
重庆交通大学学报（自然科学版）

重庆交通大学学报（自然科学版） ›› 2022, Vol. 41 ›› Issue (04): 139-144.DOI: 10.3969/j.issn.1674-0696.2022.04.21

• 交通装备 • 上一篇    

基于模糊神经网络优化的永磁同步电机改进型趋近律滑模控制研究

胡启国1，王泽霖1，曹历杰2，张军1   

  1. (1. 重庆交通大学 机电与车辆工程学院，重庆 400074； 2. 川庆钻探工程公司 安全环保质量监督检测研究院，四川 广汉 618300)
  • 收稿日期:2020-03-02 修回日期:2020-08-19 出版日期:2022-04-13 发布日期:2022-04-13
  • 作者简介:胡启国（1966—），男，重庆人，教授，博士，主要从事机械系统动力学方面的研究。E-mail：swpihqg@163.com 通信作者：王泽霖（1997—），男，甘肃兰州人，硕士研究生，主要从事电机控制方面的研究。E-mail：1942027030@qq.com

Improved Approach Law Sliding Mode Control for Permanent Magnet Synchronous Motor Based on Fuzzy Neural Network Optimization

HU Qiguo1, WANG Zelin1, CAO Lijie2, ZHANG Jun1   

  1. (1.School of Mechatronics and Vehicle Engineering, Chongqing Jiaotong University, Chongqing 400074, China； 2. Institute of Safety and Environmental Quality Supervision Inspection, Chuanqing Drilling Engineering Company, Guanghan 618300, Sichuan, China)
  • Received:2020-03-02 Revised:2020-08-19 Online:2022-04-13 Published:2022-04-13

PDF

摘要： 为进一步提高永磁同步电机滑模控制调速系统性能，在传统指数趋近律滑模控制基础上，采用了一种fal函数来代替传统符号函数，设计了一种改进型指数趋近律并用李雅普诺夫函数验证了其稳定性。借助模糊神经网络对改进指数趋近律参数进行动态优化，设计出永磁同步电机滑模转速控制器。仿真分析结果表明：与传统指数趋近律相比，改进指数趋近律减小了67.1%的超调，在负载扰动时的转速下降幅度减小了22.2%，转速恢复时长缩短了0.01 s；经模糊神经网络优化的改进指数趋近律又进一步减小了50%的超调，负载扰动时的转速下降幅度再次减小了8.9%，转速恢复时长再次缩短了0.003 2 s。

关键词: 车辆工程；永磁同步电机；滑模控制；指数趋近律；fal函数；模糊神经网络

Abstract: In order to improve the performance of permanent magnet synchronous motor sliding mode control system, a fal function was used to replace the traditional symbolic function, which was based on the traditional exponential approach law sliding mode control. And an improved exponential approach law was designed, and its stability was verified by Lyapunov function. Then, the parameters of the improved exponential approach law were dynamically optimized by fuzzy neural network, and the sliding mode speed controller of permanent magnet synchronous motor was designed. Through simulation analysis, the results show that: compared with the traditional exponential approach law, the improved exponential approach law reduces the overshoot by 67.1%; the speed drop amplitude under load disturbance is reduced by 22.2%, and the speed recovery time is shortened by 0.01s. The improved exponential approach law optimized by the fuzzy neural network further reduces the overshoot by 50%; the speed drop amplitude under load disturbance is reduced by 8.9% again, and the speed recovery time is shortened by 0.0032s again.

Key words: vehicle engineering; permanent magnet synchronous motor; sliding mode control; exponential approach law; fal function; fuzzy neural network

中图分类号: