A multiobjective ranking based finite states model predictive control scheme applied to a direct matrix converter

Matrix converters are capable of managing bidirectional power flow while fully controlling the source currents and load voltages. The topology is characterized by its reduced size and weight compared to Voltage Source Inverters because there is no DC link stage. Recently, diverse control methods have been proposed for the direct matrix converter. In particular, the finite states model predictive control (FS-MPC) gives a valid alternative for the control of this converter. However, the determination of the control parameters or weighting factors used in the traditional schemes has not been possible to treat systematically. In this work, a novel method for searching and determining the optimum state to apply is proposed that does not require weighting factors. This method uses ideas from multiobjective optimization theory and is based on obtaining a ranking of each possible solution respect to the rest of the solutions, eliminating the use of weighting factors. Simulation results are presented to validate the proposed control scheme.