To address the high pa we conduct a study on rameter dependency of control performance in PMSM(Permanent Magnet Synchronous Motor), the DPCC(Dead-beat Predictive Current Control) system is studied,incorporated the MI(Multi-Innovation) theory into the EKF(Extended Kalman Filter) parameter identification algorithm. Simulation models of the MI-EKF(multi-innovation extended Kalman filter) with different innovation lengths and the conventional EKF are constructed. Experimental results demonstrate that the inductance and flux linkage parameters identified by the MI-EKF observation algorithm exhibit superior steady-state and dynamic performance compared to the EKF algorithm. By combining MI-EKF and DPCC to obtain accurate parameter nominal values, the issue of DPCC performance degradation caused by system uncertainty is resolved. The harmonics of rotor speed and stator current is reduced achieving better dynamic performance and robustness in the PMSM control system.

permanent magnet synchronous motor ( PMSM), multi-innovation extended Kalman filter ( MI-EKF), parameter identification, dead-beat predictive current control(DPCC)