Enhanced Position Observer Based on High Flat CIC Filter in Sensorless IPMSM Drives

Accurate rotor position information is crucial to realize excellent performance of position sensorless interior permanent magnet synchronous motor (IPMSM) control. Nevertheless, position estimation methods based on the back electromotive force (BEMF) model are often disturbed by the adverse impact of the inverter nonlinearity and the rotor flux spatial harmonics. For the purpose of improving the position observation accuracy, a high flat cascaded integrator-comb (CIC) filter embedded between sliding-mode observer (SMO) under the $\gamma \delta $ frame and phase-locked loop (PLL) is proposed, in which genetic algorithm (GA) is utilized to optimize its structural parameters with the optimization objectives of both harmonic suppression capability and group delay, ensuring that the harmonic components contained in the BEMF can be suppressed sufficiently in a wide frequency region and IPMSM has excellent dynamic performance simultaneously. On the other hand, the proposed approach adopts $\gamma \delta $ frame structure, thus avoiding the phase delay problem of the low-pass filters (LPFs). Additionally, owing to the unnecessity of frequency adaptation, the difficulty of structural design is reduced. Experimental results on a 1.5-kW IPMSM control platform demonstrate the effectiveness and feasibility of this strategy.