Finite-control-set multistep model predictive control based on incremental prediction model for SPMSM drivers

Multi-step model predictive control (MMPC) for motor drives can reduce the switching loss, retard temperature rise, and improve the reliability of the control system. However, MMPC has a high computational complexity, which requires advanced microprocessors and raises the cost. In this paper, an optimized deadbeat voltage vector strategy based on receding horizon control is proposed. The MMPC model for PMSM is established, and then the proposed strategy transforms the multi-step model predictive voltage control (MPVC) into one-step MPVC, which simplifies the calculation process. Furthermore, the prediction horizon is paid attention to, and its contribution is analysed. Simulation results verify the effectiveness of the proposed method.