Comparative Evaluation of Model Predictive Control and Field-Oriented Control for PMSM Drives Using a Three-Level NPC Inverter in High-Voltage EV Applications

Abstract. In electric vehicles (EVs), traction inverters have been widely studied from a topology perspective, modulation, and control strategies. With EV manufacturers adopting higher voltage batteries to take advantage of lower currents, higher power density, and faster charging, this trend has increased interest in multilevel inverters as potential replacements for conventional inverters. Their higher efficiency, better waveform quality, lower stress on semiconductors, and inherent fault tolerance make them attractive for this application. At the same time, advances in automotive ECUs (Electronic Control Units) have made it possible to implement more high computational control strategies, opening the door to predictive control such as Finite-Control-Set Model Predictive Control (FCS-MPC), which leverages direct selection of switching states and offers very fast dynamic response. In this study, a comparative analysis is conducted between classical FCS-MPC and the conventional Field- Oriented Control (FOC) for a PMSM drive supplied by a three-level Neutral Point Clamped (NPC) inverter. The two strategies have been compared in terms of time response and the steady-state error of both torque and speed and the Total Harmonic Distortion (THD) of the stator currents. The results of the simulations prove that FCS- MPC provides faster torque response, while FOC outperforms in terms of precision with lower current distortion. In general, the results highlight advantages and disadvantages of both predictive and conventional control strategies and help clarify which approach may be better suited for next-generation high-voltage EV traction systems.