Accelerated Model Predictive Control based on a Greedy Algorithm with Round-Robin Priority for N-level Single-Phase Flying Capacitor Converter in Solid-State Transformer

Abstract

This paper presents a control method for reducing the execution time based on model predictive control in a grid-connected single-phase N-level flying capacitor converter (FCC). The proposed method controls the grid current and flying capacitor voltage through two layers with a hierarchical structure finite-control-set model predictive control (FCS-MPC). In particular, unlike conventional model predictive control, using the greedy algorithm, the number of switching combinations is reduced by segmenting power devices into groups and sequentially determining switching states. The execution time is shortened because the reduction in the number of switching combinations is diminished. Moreover, because the maximum number of switching states to be compared increases in linear proportion to the voltage level, the execution time does not increase significantly as the voltage level is increased. As a result, the switching frequency can be increased by increasing the sampling frequency, and the total harmonic distortion (THD) of the grid current can be reduced. Evaluation of the validity of the proposed method via an experiment on a scaled-down prototype of a single-phase 2 kW 7-level FCC is also detailed herein.