微特电机 ›› 2018, Vol. 46 ›› Issue (4): 75-79.doi: 1004-7018-46-4-75-79

• 驱动控制 • 上一篇    下一篇

电流-位置神经网络模型的构建与SRM转矩控制

党选举,陈恩普,姜辉,伍锡如,彭慧敏   

  1. 桂林电子科技大学,桂林 541004
  • 收稿日期:2017-05-09 出版日期:2018-04-28 发布日期:2018-04-28
  • 作者简介:党选举(1965—),男,教授,主要从事非线性系统建模与控制、精密智能机电测控系统等方面的研究。
  • 基金资助:
    国家自然科学基金项目(61263013);国家自然科学基金项目(61603107);广西信息科学实验中心项目(20130110);广西自然科学基金项目(2016GXNSFDA380001);广西自然科学基金项目(2015GXNSFAA139297);智能综合自动化高校重点实验室基金资助项目(2016)

Torque Ripple Suppression of Switched Reluctance Motor Based on a Current-Position Neural Network Model

DANG Xuan-ju, CHEN En-pu, JIANG Hui, WU Xi-ru, PENG Hui-min   

  1. Guilin University of Electronic Technology,Guilin 541004,China
  • Received:2017-05-09 Online:2018-04-28 Published:2018-04-28

摘要:

常用的基于转矩分配函数(TSF)的控制方法从转矩-电流非线性关系得到参考相电流,数学关系复杂,参数难以获得,不易于工程实现。根据相电流平方之和与转子位置角所表现出的特殊关系,提出了基于电流-位置神经网络模型的相电流平方之和计算方法,直接由转子位置角计算相电流平方之和,再通过分配函数得到参考相电流。仿真结果表明,该方案能有效降低开关磁阻电动机转矩脉动。

关键词: 开关磁阻电动机, 转矩脉动, 电流-位置神经网络模型, 转矩分配函数

Abstract:

The conventional control method based on the torque sharing function (TSF) is difficult to apply to engineering, because the reference phase currents are obtained from the complex torque-current relationship. According to the special relationship between the sum of phase currents square and the rotor position angle, a sum of phase currents square calculation method based on current-position neural network model was proposed. The sum of phase currents square was calculated from the rotor position angle directly, then reference phase currents were obtained by the sharing function. The simulation results verify that the proposed control strategy can reduce SRM torque ripple effectively.

Key words: switched reluctance motor (SRM), torque ripple, current-position neural network model, torque sharing function (TSF)

中图分类号: