New approach to EKF-based sensorless control using parallel structure for non-salient pole permanent magnet synchronous motors

This paper proposes a new sensorless control scheme of non-salient pole permanent magnet synchronous motors (PMSMs). To solve the drawback of Extend Kalman Filter (EKF) needing long computation time, a new sensorless scheme is achieved by using a parallel structure of EKF model, which alternatively operates at the sampling period. The proposed scheme can greatly save computation time with a similar performance compared with full-order EKF. Therefore, it can overcome the low computation ability of low cost digital processors or have extra time for precise control algorithm. Experimental results validate the theoretical analysis and feasibility of the proposed estimation technique. In addition, the computation time of the proposed reduced-order EKF algorithm using a parallel structure has been compared with that of a full-order EKF algorithm.