PMSM Inductance Surface Adaptive Identification Decoupled From Rotor Position Error and Speed Coupling Effect

The conventional online inductance estimation of permanent magnet synchronous motor (PMSM) generally rely on the accurate rotor position. Meanwhile, the inductance surface forming requires the scanning of operation conditions. In order to improve the universality and accuracy, the online inductance estimation method based on the differential rotation axis high-frequency signal injection (DRHSI) is proposed, which is immune to rotor position error. Furthermore, the coupling effect between the high-frequency (HF) estimation model and PMSM rotation speed is analyzed, which causes more than 15% inductance estimation error at two times of rated speed. A speed coupling effect compensation method is introduced to improve the estimation accuracy. Then, the inductance surface forming method based on the forgetting factor self-adapted recursive least squares (FFSA-RLSs) is studied. The inductance surface can be achieved by at least three data points with forming error less than 5.9%. and through the iteratively optimization of FFSA-RLS, the forming error of inductance surface can be further reduced to 3.1%. The effectiveness of the proposed method is verified on a 2.2 kW PMSM platform.