Modeling and Simulation of a DC Micro-Grid with a Model Predictive Controller

Abstract

This paper presents the modeling strategies of a micro-grid to control the power supply of a battery bank by adopting a Model Predictive Controller (MPC). The grid was sized to light two tennis courts at a university sports complex, which is not connected to the national power grid and thus must be a stand-alone setup. The paper starts with an introduction that defines the statement of purpose and the state of the art. Then continue with the power generators and storage modeling: photovoltaic (PV) modules, wind turbine, buck converter (takes power from the rectified national grid) and the battery bank. The MPC was designed to effectively manage the energy supplied to the batteries, depending on the state of charge, hence the controller output is the signal used to regulate the charging current. The data used for prediction is the meteorological measures taken during three years using an in-situ weather station that collected irradiance, wind speed and direction, temperature, and pressure. Finally, as the entire control system was simulated step by step using MATLAB/Simulink, the components and systems behavior graphs are shown to lead to analysis and conclusion remarks.