Position Estimation Accuracy Improvement for SPMSM Sensorless Drives by Adaptive Complex-Coefficient Filter and DPLL

This paper presents an improved high-frequency pulsating square-wave signal injection-based sensorless control for surface-mounted permanent-magnet synchronous machines (SPMSM). The use of low pass filter in the traditional sensorless method results in phase delay and, consequently, degrades the dynamic behavior. In order to eliminate such drawback, the proposed method adopts an adaptive complex-coefficient filter (ACCF) that does not cause phase delay and the use of compensation technique is avoided. Moreover, a dual-phase-locked loop (DPLL) is utilized to extract the accurate rotor position and eliminates the position estimation error in dynamic state. The small signal model of ACCF is build, and then the unified parameter selection guideline for both ACCF and DPLL is given in terms of comprehensive transfer function analysis. The proposed method has been carried out by both simulation and experiment, where a Higale platform with a 0.2-kW SPMSM is adopted to verify the performance of proposed method.