Improved Model Predictive Control for Hybrid ANPC with Reduced Computational Burden

This paper proposes an enhanced finite-control-set model predictive control (FCS-MPC) for a three-level Si/SiC hybrid active neutral point clamped (3L-HANPC) converter. The novel MPC strategy optimally manages multiple objectives by leveraging all control degrees of freedom, a departure from the conventional MPC approach. Using the suggested cost function eliminates the need to calculate duty cycles for each vector, thereby minimizing division operations. Simultaneously, stable switching frequency and reduced neutral-point (NP) voltage fluctuation are achieved by selecting voltage vector sequences that meet the switching frequency requirements of low-frequency cells (LFC) and high-frequency cells (HFC). Experimental and simulation results validate the effectiveness of the proposed control approach.