Online Multi-Parameter Observation of IPM Machine With Reconstructed Nonlinear Small-Signal Model Based on Dual EKF

Flux nonlinearity due to cross-coupling saturation is critical for interior permanent magnet (IPM) machine control. To obtain accurate parameters in real-time, this article proposes a novel reconstructed nonlinear small-signal model which simplifies the observability analysis and design of the observer. Compared with traditional methods, the observability of the reconstructed small-signal model shows that all resistance and inductance parameters are observable under entire operation region. The flux cross-coupling saturation is also well considered. A dual-extended-Kalman-filter based observer for the reconstructed model is introduced to achieve fast parameters convergence. The effectiveness of the proposed observer is experimentally evaluated with a 60 kW IPM machine and verified by finite-element analysis.