Simplified Model Predictive Control for Permanent Magnet Synchronous Machine based Back-to-Back Converter

Abstract

To decide the most appropriate switching vector, traditional model predictive control (MPC) algorithms have to suffer extreme time burden for 8-time searching in each sampling period. Meanwhile, the traditional MPC methods have some other demerits, such as complex cost function to get reasonable good performance. In order to solve such defects, one new model predictive control method based switching table (MPC-ST) is proposed in this paper. Firstly, the MPC-ST and MPC are applied to permanent magnet synchronous machine (PMSM) drive system based back-to-back converter. Secondly, in the same conditions, a detail comparison between MPC-ST and MPC is conducted by comprehensive simulation. It fully demonstrates that the new MPC-ST considering other performance parameters can simplify cost function greatly.