Low Complexity Model Predictive Flux Control Based on Discrete Space Vector Modulation and Optimal Switching Sequence for Induction Motors

Abstract

In this paper, a low-computational burden model predictive flux control (MPFC) based on discrete space vector modulation (DSVM) and the optimal switching sequence (OSS) is proposed for achieving switching frequency (SF) and computational burden efficiencies in motor drives fed by two-level voltage source inverter. The DSVM is used to extend the prediction candidates of MPFC and greatly improve the performance of the controller. A generalized minimum flux error method independent of the number of virtual vectors is derived to cancel the exhaustive optimization method and lower the execution time of the proposed algorithm. In addition, new over-modulation and OSS schemes are designed to optimize the use of DC-link voltage and mitigate the inverter SF when implementing the optimal control action into switching states. The comparative experimental results show that without significant performance degradation, the proposed strategy provided about 50% SF and 25% execution time reductions compared to the classic MPFC methods.