Sensor-Optimized FCS-MPC Strategy for a Three-Phase PUC5 Inverter with CMV Mitigation and Capacitor Voltage Balancing

This paper proposes a reduced-sensor modified Finite Control Set Model Predictive Control (FCS-MPC) strategy tailored for a three-phase 5-level Packed U-Cell (PUC) inverter, with an emphasis on common-mode voltage (CMV) reduction and DC-link capacitor voltage balancing. The proposed approach leverages an Extended Kalman Filter (EKF) to estimate phase currents, allowing the control system to operate with only a single current sensor. This reduced-sensor strategy addresses the limitations of conventional current measurement methods by predicting unmeasured phase currents based on model-based prediction and real-time feedback. By integrating the EKF, the control method not only minimizes reliance on multiple sensors but also enhances robustness against measurement noise. Simulation results validate the effectiveness of the proposed method, demonstrating low current Total Harmonic Distortion (THD), nearly zero CMV, and well-balanced capacitor voltages. The proposed reduced-sensor FCS-MPC approach simplifies the hardware setup while maintaining high reliability and improved power quality, as verified by MATLAB/Simulink simulations.