基于遗忘因子的高速永磁同步电机在线参数辨识自适应控制研究
Adaptive control research on online parameter identification of high-speed permanent magnet synchronous motors based on forgetting factor
摘要
针对高速永磁同步电机在运行过程中因参数时变引发的控制性能退化问题,本文提出一种基于多参数在线辨识的自适应控制策略.该策略通过构建模型参考自适应系统(MRAS)与递推最小二乘法(RLS)融合的混合参数辨识框架,实现了对定子电阻、d 轴和 q 轴电感、永磁体磁链及转动惯量等关键参数的实时高精度估计;并进一步依据辨识结果动态调节转速环与电流环控制器的参数.为进一步提升参数辨识的收敛速度与估计精度,本文提出一种递推最小二乘法的遗忘因子自适应律.实验验证表明,相较于传统控制策略,当负载工况与电机参数发生显著变化时,本文所提方法能够实现电机参数的快速精准辨识,并有效抑制因参数失配引发的转速波动与转矩脉动问题,显著提升了电机在时变工况下的控制鲁棒性.
To address the control performance degradation of high-speed permanent magnet synchronous motor(HSPMSM)caused by time-varying parameters during operation,this paper proposes an adaptive control strategy based on online multi-parameter identification.The strategy establishes a hybrid parameter identification framework that integrates model reference adaptive system(MRAS)with recursive least squares(RLS),enabling real-time high-precision estimation of critical parameters including stator resistance,d/q-axis inductances,and permanent magnet flux linkage.Based on the identification results,the parameters of both speed and current loop controllers are dynamically adjusted.To further enhance the convergence speed and estimation accuracy of parameter identifi-cation,an adaptive forgetting factor mechanism for RLS is proposed.Experimental results demonstrate that com-pared with conventional control strategies,when significant changes occur in load conditions and motor parameters,the proposed method achieves rapid and accurate motor parameter identification while effectively suppressing speed fluctuations and torque ripples induced by parameter mismatch,thereby significantly improving the control robust-ness under time-varying operating conditions.
年,卷(期):
2026, 45(9)
分类号:
TM352
机标分类号:
TM351
TP273
TM76
在线出版日期:
2026年10月14日
基金项目:
52207222:国家自然科学基金,