Model Predictive Control with Reduced Computation for N-cell Cascaded Flying Capacitor H-Bridge Converter in Solid-State Transformer

Abstract

This paper describes research conducted on single-phase N-cell cascaded flying capacitor H-bridge (CFCHB) multilevel converter in the solid-state transformer (SST). The proposed algorithm has a hierarchical structure based on finite-control-set model predictive control (FCS-MPC) and each layer has a different control goal. Thus, the cost functions consist of a single-objective control set and are separated by each layer. It reduces the number of candidates to be compared, and it can lead to reducing the computational burden. Therefore, in the proposed method, the sampling period can be shorter than the conventional method, so the sampling frequency can increases. The effectiveness of the proposed method was verified by a simulation using PSIM on a single-phase 9kW 3-cell CFCHB.