An Improved Model Predictive Voltage Control With Reduced Computational Burden for T-Type Three-Phase Three-Level Inverters

Abstract

The traditional finite-set model predictive voltage control for three-phase three-level voltage source inverters with LC-filter suffers from high computational burden, fussy weighting factor adjustment, and variable switching frequency. To address these issues, this article proposes an improved model predictive voltage control scheme with optimal switching sequence. Two voltage vector switching modes are designed to eliminate weighting factors and achieve neutral-point voltage balancing. Furthermore, a subsector optimization method based on boundary conditions is proposed, which avoids the calculation of multiple groups of duty cycle and hence, greatly reducing the computational burden. Additionally, the proposed method operates at a constant switching frequency. The effectiveness of the proposed algorithm is verified by experimental results based on a laboratory prototype from various perspectives, including steady-state and dynamic performance, algorithm execution time, and parameter sensitivity.