A Forecast-Driven Energy Management Methodology for Microgrids under Seasonal Variability

This paper introduces a dynamic energy management strategy based on model predictive control (MPC) for a microgrid with photovoltaic generation and an energy storage system. The strategy is designed to be able to acclimate to different seasonal situations. Moreover, to model realistic scenarios, some forecasting uncertainties are made current, using noise models for photovoltaic, load and cost variables. The performance of the MPC was also compared with an open-loop controller with pre-known trajectory prediction. Moreover, some week scenarios are carried out using real and processed solar generation data and load profiles from Milan, Italy, considering summer and winter periods. The results show the MPC strategy is able to optimize grid transactions, while the stability of the battery is kept even in hard and noisy scenarios.