Predictive Current Control With Input Filter Resonance Mitigation for a Direct Matrix Converter

Abstract

This paper presents a control method for direct matrix converters that combines the advantages of predictive control with active damping. The active damping is used to suppress the resonance of the input filter. A discrete-time model of the converter, the input filter, and the load parameters are used to predict the behavior of the input reactive power on the supply side and the output currents for each valid switching state. The control method selects the best commutation state, according to an optimizing algorithm and a cost function, in order to generate unity input displacement power factor with high-quality waveforms. The active damping method is based on a virtual harmonic resistor that damps the filter resonance. This paper shows simulated and experimental results to demonstrate that the proposed control method can generate good tracking of the output current references, achieve unity input displacement power factor, and reduce the input current distortion caused by the input filter resonance.