Sensorless Position Estimation Method of Winding-Segmented LPMSM Based on a Composite MRAS

This paper proposes a composite model reference adaptive system (MRAS) method for the electromagnetic parameter position-sensitive problem caused by the primary segmented structure of winding-segmented linear permanent magnet synchronous motor (WS-LPMSM). Due to the segmented winding structure, end effects between adjacent stators of ws-LPMSM can cause nonlinear changes in key parameters such as inductance and magnetic flux with the mover position, leading to the failure of traditional sensorless control. In this paper, by constructing the composite reference model of two adjacent segments of the motor, the time-varying influence of parameters in the segment transition region is eliminated and a cooperative control strategy is designed to suppress the asymmetry between segments. The experimental results show that the position estimation error of this method is reduced by about 34% compared with the traditional MRAS.