Sensorless speed estimation of PMSM using a hybrid method

A hybrid estimator which combined the advantages of sliding mode estimator and extended Kalman filter for sensorless control of PMSM was proposed. The algorithm is more effective than sliding mode estimator for a motor with a distorted back-emf distribution. The hybrid estimator established a three-order EKF, considering the high order harmonics components of back-emf as system and measurement noise. It is easier to tune the parameters for the order lowered EKF estimator than traditional EKF estimator. It has a good performance for PMSM with no sinusoidal back EMF distributing while sliding mode estimator has a worse capability. Compared with the sliding mode estimator, traditional EKF estimator and hybrid estimator, the advantages and disadvantages of the hybrid estimator are illustrated.