Domestic microgrid energy management: Model Predictive Control strategies experimental validation

The growing interest in meeting “20-20-20” European commitment has led towards enhancing renewable energy source deployment and fossil fuel use reduction. Distributed generation systems and thus microgrids represent an important research focus for their capability to allow of high penetration of renewables in the land as well as reducing carbon footprint. It is well known that microgrid performances and reliability strongly depend on generators and load interaction, highlighting the importance in developing a proper energy management strategy. A Model Predictive Control logic has been already applied to a model of a domestic microgrid composed of PV panels, Fuel Cell (FC) and a battery pack. This work presents results obtained with an experimental microgrid test bench developed to validate results previously obtained only numerically. Two controller levels have been developed to that aim, to deal both with low level component management and high level supervision by the MPC controller.. Results showed a stable and reliable operation of the microgrid, referring to two main strategies (SOC follow and FC follow), having different characteristics in terms of battery and fuel cell behavior over time.