基于仿射算法的永磁同步电机参数在线辨识

To decouple coupled parameters caused by rank-deficient identification models, an adaptive affine algorithm is proposed. By injecting transient weak-field current on the d-axis, it achieves online decoupled identification of stator resistance, inductance, and PM flux linkage for surface-mounted PMSMs. A nonlinear voltage-drop-immune model is built considering inverter dead-time, switching delays, and diode saturation voltage. Experiments verify improved accuracy and faster convergence.

永磁同步电机参数在线辨识时由于辨识模型非满秩而存在耦合,为了实现参数解耦辨识,提出1种自适应仿射参数辨识算法,通过d轴注入瞬时弱磁电流信号,实现表贴式永磁同步电机定子电阻、电感及转子永磁体磁链参数在线解耦辨识.考虑逆变器死区时间、开关管开通关断延时及续流二极管饱和压降等产生的非线性电压降,构建不受非线性电压降影响的参数辨识模型.实验结果表明,所提方法能够提高电阻、电感及转子磁链的辨识精度,同时加快参数在线辨识收敛速度.