Multistep Predictive Current Control for Electrical Drives With Adaptive Horizons

Since the number of candidate vectors is increased in the multistep model predictive control (MMPC) strategy, its calculation burden becomes a serious problem in implementation. To make full use of the limited processor resources effectively and improve control performances, an MMPC with the adaptive prediction horizon (PH), control horizon (CH), and online selecting weighting factor is proposed in this article and applied to a permanent magnet synchronous motor (PMSM) driving system as the current controller. A hysteresis principle is designed to judge the current operating state to online adjust PH and CH in the MMPC within the normal timescale, and a real-time branch and bound algorithm is designed to online select a suitable weighting factor based on the overtimescale caused by the long CH. More effective availability of the processor resources by the proposed method is analyzed and its advantages including improved current impact, reduced switching frequency, and decreased influences of some parameter mismatches with suitable stator current quality are demonstrated by the simulation and experimental results compared with some conventional control strategies.