华中科技大学学报(自然科学版)2025,Vol.53Issue(1) :14-20.DOI:10.13245/j.hust.250020

基于交直轴电感关联和机械参数的PMSM参数辨识

PMSM parameter identification based on inductive correlation and mechanical parameters

杨路鹏 韩瑜 丁北辰 冯国栋
华中科技大学学报(自然科学版)2025,Vol.53Issue(1) :14-20.DOI:10.13245/j.hust.250020

基于交直轴电感关联和机械参数的PMSM参数辨识

PMSM parameter identification based on inductive correlation and mechanical parameters

杨路鹏 1韩瑜 2丁北辰 3冯国栋1
扫码查看

作者信息

  • 1. 中山大学智能工程学院,广东 深圳 518107
  • 2. 中山大学智能工程学院,广东 深圳 518107;广东省消防科学与智能应急技术重点实验室,广东 广州 510006
  • 3. 中山大学先进制造学院,广东 深圳 518107
  • 折叠

摘要

针对永磁同步电机控制中复杂的非线性多参数问题,提出了一种考虑机械参数和交直轴电感相关性的多参数精确辨识方法.首先,根据三相定子坐标系下的电感关系,建立交直轴电感的关联模型,结合电机电压、机械模型,估计阻尼系数和直流偏置导致的转矩误差.然后,建立考虑机械参数的相关电感非线性估计模型并利用最小二乘辨识参数.该方法考虑电感相关性并结合了机械模型,提高了永磁同步电机多参数辨识的精度.最后,通过实验验证和对比验证了提出方法的有效性.

Abstract

A multi-parameter precise identification method considering the correlation of mechanical parameters and mutual inductance of the dq-axis was proposed to address the complex nonlinear multi-parameter problems in the control of permanent magnet synchronous motors.Firstly,based on the inductance relationship in the three-phase stator coordinate system,a correlated model of the mutual inductance of the dq-axis was established.Combining with the motor voltage and mechanical model,the damping coefficient and torque error caused by the DC bias were estimated.Secondly,a nonlinear estimation model of the correlated inductance considering the mechanical parameters was established,and the parameters were identified using the least squares method.The proposed method considers the correlation between inductances and combines with the mechanical model,improving the accuracy of multi-parameter identification for permanent magnet synchronous motors.Finally,the effectiveness of the proposed method was verified through experimental validation and method comparison.

关键词

磁饱和/机械参数/相关电感模型/多参数/最小二乘法

Key words

magnetic saturation/mechanical parameters/correlated inductance model/multi-parameter/least square method

引用本文复制引用

出版年

2025
华中科技大学学报(自然科学版)
华中科技大学

华中科技大学学报(自然科学版)

CSTPCDCSCD北大核心
影响因子:0.813
ISSN:1671-4512
浏览量1
段落导航相关论文