Application of fuzzy decision making to the switching state selection in the predictive control of a Direct Matrix Converter

The finite states model predictive control is a promising alternative in the field of the control of power converters. Its characteristics allow simple and flexible control schemes with fast dynamics. However, the standard formulation of this type of controllers, which is based on an aggregated cost function, requires suitable weighting factors for an adequate switching state selection. The specification of these factors depend on various parameters making it a non trivial process. In this work, an alternative is proposed that replaces the standard aggregated cost function with a fuzzy decision making strategy. This strategy retains the multiple attribute nature of the state selection. As a result, the possibility of a more natural and higher level design approach to the state selection process is opened. To introduce the technique, only objectives with equal importance are considered. Simulation results are presented to illustrate and validate the approach using the control of output and input currents in the Direct Matrix Converter.