Adaptive MPC Strategy for MMCs: Bridging Computational Efficiency and System Robustness

This paper presents a novel Level Model Predictive Control (MPC) framework tailored for Modular Multilevel Converters (MMCs). The proposed method is designed to address key challenges such as computational complexity, energy balance, and harmonic mitigation while maintaining robust performance. The Level MPC leverages a dynamic prediction horizon and a holistic cost function, achieving optimal control actions with reduced computational overhead without compromising control performance. Extensive simulations validate the controller's ability to maintain energy balance, suppress circulating currents, and minimise Total Harmonic Distortion (THD) under various operating scenarios, including parameter mismatches and harmonic disturbances in the AC grid. Furthermore, the proposed method is compatible with existing control platforms, requiring no specialised hardware upgrades, thus providing a cost-effective and practical solution for high-performance MMC applications.