Comparisons of Sensorless Control Performance of Surface Permanent Magnet Synchronous Motors Based on Flux Linkage Observers

Flux linkage observers (FLOs) are widely adopted for sensorless control of permanent magnet synchronous motors (PMSMs), especially for fast and direct closed-loop startup and low-speed applications. Nevertheless, the performances for different FLOs have not been systematically evaluated. This study experimentally analyzes four typical flux observers: the Active Flux (AF) method, the Ortega method, the regression model (RM) method, and an improved Ortega method. Several key performances include dynamic response, estimation accuracy, anti-voltage offset, low-speed operation, acceleration/deceleration, speed reversal, and parameter robustness, etc., have been fairly compared. Experimental results reveal that the improved Ortega method achieves the lowest position error during low-speed and speed reversal operations compared to the other approaches. The AF method demonstrates strong immunity to voltage offsets. Meanwhile, the RM and the improved Ortega methods exhibit strong robustness against stator resistance variations. All FLOs show similar sensitivity to inductance mismatches in terms of position estimation errors. This analysis provides practical guidelines for selecting flux linkage-based sensorless strategies PMSM drives.