Adaptive Fault-Tolerant Control Based on Model-Free Predictive Current Control for PMSM With Dual Star-Delta Windings

This paper proposes a model-free predictive current control combined with an adaptive fault-tolerant control (AFTC-MFPCC) strategy for dual star-delta windings permanent magnet synchronous motor (DSDW-PMSM) under different phase shift angles. The proposed strategy enhances the operational performance of this motor when current sensor fault or open-phase fault occur in any star-connected winding phase, showing strong adaptability and robustness. Firstly, the invariance of the fundamental magnetomotive force (MMF) under single open-phase fault in DSDW-PMSM with different phase shift angles is analytically derived. Secondly, an adaptive fault-tolerant control strategy is proposed. Upon detecting a persistent zero current in any phase, it switches the adaptive fault-tolerant transformation matrix without explicit fault diagnosis, thereby avoiding the impact of current sensor faults and reducing the fault detection time. Subsequently, the xy-axis reference values are modified through current feedback, enabling the motor to adaptively transition from the fault state to the normal operation state. Furthermore, model-free predictive current control is employed to replace traditional PI control or deadbeat predictive current control in the current loop module of the vector control structure, which outputs four voltage vectors. This method avoids current tracking errors and severe waveform distortion caused by inaccuracies in motor parameter inputs. Finally, tests are conducted on a 24-slot/10-pole DSDW-PMSM with a 15° phase shift to validate the feasibility of the proposed method.