Development of Model Predictive Control for Three-Stage MMC-based Solid State Transformer to Interconnect High and Low Voltage AC Grids

The growing need for flexible and efficient power conversion has driven interest in solid-state transformers (SSTs) as a potential alternative to conventional transformers. This paper investigates a three-stage modular multilevel converter (MMC)-based SST designed for interfacing AC grids operating at different voltage levels while seamlessly integrating solar photovoltaic (SPV), battery energy storage (BES), and electric vehicle (EV) charging units. This paper further contributes a finite control-set model predictive control (FCS-MPC) scheme for the three-stage SST. Mathematical modeling of the SST is carried out, and prediction models for current and voltage are developed for each stage. Multi-variable cost functions are employed to directly generate switching pulses for the converters of the SST by incorporating system constraints directly into the optimization process. This approach eliminates the need for meticulous tuning, pulse width modulators, and reduces switching frequency. Furthermore, this paper also examines the AC fault ride-through operation of the SST during AC faults at high-voltage and low-voltage grids. In addition, a power management strategy is also investigated in this study to ensure optimal power distribution, efficient SPV and BES utilization, and dynamic local load management. A series of experiments is conducted on the hardware-in-loop platform to validate the advanced functionalities of the proposed FCS-MPC. The effectiveness of the proposed FCS-MPC against conventional control technique under both steady-state and dynamic conditions is presented.