Fast Response Finite Control Set Model Predictive Control for High-Frequency Link Matrix Converter With Virtual Space Vectors

High-frequency link matrix converter (HFLMC) has the advantages of high efficiency, high power density and high reliability, and is suitable for being used as electric vehicle charger. However, the coupling of input side and output side of the topology faces complex parameter tuning and insufficient dynamic response ability under conventional PI control strategy. To deal with this problem, fast response finite control set model predictive control (FCS-MPC) strategy is proposed, which omits the modulation part and directly selects the vector to participate in the control. The strategy is very suitable for the HFLMC because it can control multiple targets simultaneously. This control strategy has the disadvantage of poor control performance at low sampling frequency, so virtual space vectors are proposed. However, too many virtual space vectors will increase the computational burden of the controller. In this paper, sectors are divided according to the magnitude of three-phase voltage, and only the vector that makes the voltage positive is selected in each sector for calculation. This reduces the computational burden. The simulation results indicate that the fast response FCS-MPC strategy has better dynamic response performance when it is compared with PI control strategy.