Sensorless Control of Dual Three-Phase PMSM Using a Subspace Decoupling-Based Fuzzy Super-Twisting Sliding Mode Observer

To address the chattering, parameter sensitivity, and limited operating-condition adaptability of conventional sliding mode observers (SMOs) in sensorless control of dual three-phase permanent magnet synchronous motors (DTP-PMSMs), this paper proposes a subspace-decoupling-based integrated method combining a fuzzy super-twisting sliding mode observer (F-STSMO) and grey wolf optimization (GWO). In the proposed method, fuzzy control is integrated with the super-twisting sliding mode observer (STSMO) to construct a fuzzy time-varying-gain observer. The stability of the observer under time-varying disturbances is demonstrated, which effectively suppresses chattering and improves the observation performance over a wide dynamic range. By exploiting the multiple degrees of freedom of DTP-PMSMs, a GWO-based parameter estimation algorithm is developed. The sensitivity to initial settings and the convergence conditions are discussed, enabling high-accuracy real-time motor parameter estimation and enhancing the robustness of the STSMO against parameter variations. Furthermore, the Lyapunov stability constraints of the integrated F-STSMO and GWO method are analyzed. Experimental results show that the proposed integrated F-STSMO and GWO sensorless control method achieves a resistance estimation error of 0.9% for the DTP-PMSM, reduces the rotor position error caused by parameter variations by 88.6%, and keeps the rotor position error fluctuation within 2.51° under both steady-state and dynamic operating conditions.