Improved Stator Flux Linkage Observer of PMaSynRM Based on Integration Error Estimation

This paper presents an improved flux observer for sensorless control of permanent magnet-assisted synchronous reluctance motors (PMaSynRM), addressing the critical challenge of integration error accumulation in stator flux estimation. Traditional frequency-domain approaches using high-pass filters often distort flux amplitude and phase characteristics. The proposed method innovatively employs a time-domain linear state observer to directly estimate and dynamically compensate for integration errors without requiring precise motor parameters. Key contributions include enhanced immunity to DC offsets, reduced parameter sensitivity, and improved transient performance. Simulation studies under 300 rpm operation with artificial DC current disturbances demonstrate complete DC offset elimination and accurate rotor position estimation, even during 50%–100% load transitions. The observer maintains less than 0.5° position error, validating its effectiveness for medium-high speed sensorless control applications while avoiding additional losses from high-frequency injection methods.