Sensorless Control of PMSM Based on a Novel Adaptive Gain Fast Super-Twisting Sliding Mode Observer

To address the issues of insufficient speed estimation accuracy and high-frequency chattering in sensorless sliding mode control of permanent magnet synchronous motors (PMSMs), this paper proposes a novel Adaptive-Gain Fast Super-Twisting Sliding Mode Observer (AG-FST-SMO). Firstly, a new adaptive gain coefficient is designed to accommodate dynamic variations in both motor speed and load torque, while reconstructing the super-twisting sliding mode observer structure through the introduction of linear terms. The proposed AG-FST-SMO significantly reduces speed chattering while maintaining high estimation accuracy across wide speed ranges and under substantial load torque fluctuations. Furthermore, the global boundedness condition of the fast super-twisting algorithm (FSTA) is employed to verify that the designed adaptive gain coefficient satisfies stability requirements. Finally, simulation results demonstrate the effectiveness of the proposed method.