An Optimized Model Predictive Control Method for Hybrid ANPC With Fixed Switching Frequency

Abstract

This study introduces an enhanced finite-control-set model predictive control (FCS-MPC) strategy for a three-level Si/SiC hybrid active neutral-point-clamped (3L-HANPC) converter designed to optimize control performance across multiple objectives. Unlike conventional model predictive control (MPC) schemes, the proposed approach fully exploits available control degrees of freedom, offering significant improvements in both efficiency and computational simplicity. By reformulating current constraints into voltage constraints, the method reduces computational demands, specifically by minimizing division operations; furthermore, tailored voltage vector sequences are selected to satisfy the switching frequency requirements of both low-frequency cells (LFCs) and high-frequency cells (HFCs), enabling fixed switching frequencies and mitigating neutral-point (NP) voltage fluctuations. This optimized control framework also promotes a more uniform loss distribution and enhanced thermal management, which is critical for high-power drive applications. Experimental validation using a 4-SiC hybrid active neutral point clamped (HANPC) inverter platform demonstrates the superior steady-state and dynamic performance of the proposed FCS-MPC while significantly reducing the computational burden on the digital signal processor (DSP), underscoring its practical feasibility for real-world applications.