Constant switching frequency model predictive control of MMC-PET for grid integrated solar photovoltaic system

Abstract

This paper introduces a novel control technique for the modular multilevel converter-based power electronic transformer (MMC-PET). MMC-PET under consideration encompasses an integrated solar photovoltaic (SPV) system, an MMC-based isolated DC–DC converter (MIDC), and a three-phase MMC interconnected with a power grid. A constant switching frequency model predictive control (CSF-MPC) is developed for the grid-tied MMC of the MMC-PET. The developed CSF-MPC is designed to efficiently handle multiple control objectives of the MMC-PET, including active power injection, common DC-link voltage regulation, harmonic current compensation, and low voltage ride-through (LVRT) support. Further, to ensure effective control of the MMC-PET during both LVRT and normal operating modes, a seamless transition control technique is developed. This developed technique enables the MIDC to operate the SPV system at its maximum power point under normal conditions whereas during LVRT mode, dynamically shifts its control strategy to regulate the common DC-link voltage and limit solar power generation. In addition to this, an arm current sensorless voltage balance control technique is implemented to regulate submodule capacitor voltages. The developed control technique offers benefits such as a predictive nature, reduced tuning efforts, minimized sensing variables, and constant-frequency switching pulses. Additionally, it enhances transient response and adaptability to varying operating conditions. Simulation and experimental results validate the performance of the developed control method in various operating scenarios, such as variable solar irradiation, linear and non-linear load, unbalanced load conditions, and LVRT events.