Robust Sensorless Control of Synchronous Reluctance Motor Based on PDSH-KF Considering Dynamic Magnetic Saturation Effect

The accuracy of position estimation for synchronous reluctance motor (SynRM) sensorless control suffers from the magnetic saturation effect and uncertain system noise, especially in the conditions of dynamic, low-speed operation, and parameter mismatch. In this article, a robust sensorless control strategy built on a positive definite Sage-Husa Kalman filter (PDSH-KF) considering the dynamic magnetic saturation effect is proposed. First, the static inductance and dynamic inductance caused by magnetic saturation are included in the establishment of the extended back EMF (EEMF) model. Furthermore, relying on the Kalman filter (KF), an EEMF estimator based on PDSHKF is proposed to track the position of SynRM, in which the mean and covariance of the system noise can be adjusted online. Moreover, the observer divergence caused by the negative definite noise covariance matrix is avoided. In this way, the EEMF can be estimated accurately under varying operating conditions. Finally, the operating-frequency related EEMF component is extracted in no phase lag way by second-order generalized integration. The effectiveness of the proposed sensorless control scheme is verified through an experimental prototype.