A Robust Encoderless Control Method for PMSM Based on An Improved Extended Kalman Filter

Model-based encoderless control methods typically require precise flux linkage or back-EMF data, which is commonly obtained through various observation techniques. However, their performance is often degraded by system disturbances and measurement noise. In this paper, we introduces an innovative encoderless control framework utilizing an improved extended Kalman filter (EKF). We developed a new flux model for the Kalman filter based on active flux concept to prevent incorrect convergence. Additionally, the system dynamically self-corrects using stator current estimation errors, enhancing robustness against parameter mismatch.