An Improved Model Parameter‐Free Predictive Control With Fast Gradient Updating for Three‐phase Three‐Level Inverters

ABSTRACT Look‐up‐table (LUT) based model parameter‐free predictive control (MFPC) suffers from long update time and large storage requirement, particularly when applied to multi‐level inverters. To address these challenges, this paper proposes an improved MFPC with fast gradient updating MFPC. First, an improved LUT is developed by analysing the α and β components of the voltage vectors. Compared to the traditional LUT, the improved version enhances storage efficiency and gradient update speed by approximately 63.3 %. Second, a fast optimization method based on the improved LUT is introduced, in which the cost function of at most two subsectors needs to be evaluated to complete the optimization. Finally, to eliminate the weighting factors, two different switching sequences are designed for each subsector based on redundant small voltage vectors, realizing the automatic balancing of the neutral point voltage. Extensive theoretical analysis and experimental results validate the effectiveness of the proposed method.