Novel Model Predictive DTC for Three-to-Five Phase Direct Matrix Converter Fed Induction Motor Drive

In terms of compactness and complexities, traditional two-stage ac-dc-ac converters used in industrial drive applications are less efficient than the direct matrix converter (DMC). As the application of multiphase drives evolves, DMC is discovered as a supply option; thus, accurate control of DMC fed multiphase drives is a technological challenge. This paper considers a case study of a three-to-five phase DMC fed five phase induction motor drive. By utilizing the virtual vector concept, a model predictive direct torque control (MPDTC) scheme with lower computational burden is developed for such drive scheme. To eliminate the effect of the output voltage vector's xy components, a volt-second balancing is established. The rank analysis method is then used to select the optimum voltage virtual vector for accurately controlling the speed and torque of the FPIM. Furthermore, DMC's input grid power factor is kept constant at unity. A simulation study of DMC fed FPIM is provided along with hardware validation results.