Abstract

Modular Multilevel Matrix Converters (M3C) provide an efficient solution for high-power ac-to-ac conversion, offering high modularity, superior power quality, and reduced power filtering requirements. However, controlling their Circulating Currents (CCs) and Cluster Capacitor Voltages (CCVs) remains a significant challenge due to the strong coupling between system variables. To address these issues, this article presents a novel cascade-free model predictive control strategy that introduces a linearized M3C model, incorporating the Common-Mode Voltage (CMV) as an additional control input. Unlike conventional control structures relying on multiple nested loops or iterative procedures, the proposed approach computes all control actions in a single stage by solving a constrained optimization problem. Simulation and experimental results demonstrate that the proposed methodology effectively reduces CCV fluctuations and lowers the magnitude of CCs (peaks and effective values), even under constrained operating conditions. These results confirm that including the CMV as a control input provides additional degrees of freedom, making it a compelling alternative for high-performance M3C applications.