A Memory-based Model Predictive Control for Multiphase Electric Drives Using SiC Switches

Multiphase electric drives are considered an interesting solution when industrial applications require extra reliability and efficiency. To exploit the mentioned above advantages, model predictive control (MPC) strategies have been widely tested in recent years. A suitable harmonic mitigation has been obtained thanks to the multi-vector solutions for MPC techniques. However, the appearance of semiconductors based on silicon carbide (SiC) technology can change the control design constraints. Due to the reduced sampling period, the viability of multi-vector approaches might be in question. In this scenario, this work proposes the inclusion of a memory stage in an MPC scheme to mitigate the current harmonic distortion, reduce the switching frequency and compensate the system asymmetries. This alternative focuses on the past behavior of the secondary components due to the main role of this subspace in the electric drive performance. The capability of the proposed control technique has been confirmed using simulations in MATLAB/Simulink.