Model-Predictive Direct Flux Vector Control for Synchronous Reluctance Machine

This paper proposes a model-predictive direct flux vector control (MPDFVC) for synchronous reluctance (SynRel) machines. Compared with the widely adopted vector control based in rotor synchronous rotating coordinate system, the proposed control strategy is built upon control variables in stator flux-oriented frame, i.e., stator flux and torque-producing current. The integration of direct flux control with predictive control aims to enhance dynamic performance and mitigate the cross-coupling effects of SynRel machines by utilizing the inherent bandwidth advantages of predictive control. Key features of the proposed controller include torque and flux reference tracking terms, as well as a soft constraint on switching effort to reduce switching frequency and current distortions. Validation through simulation and experimental tests demonstrates the effectiveness of the proposed approach, showing fast dynamic response, reduced overshoots and undershoots of stator current, and minimized cross-coupling between state variables.