Federated Learning Enabled Prediction of Energy Consumption in Transactive Energy Communities

The prediction of the net electricity demand is crucial to the management and optimization of transactive energy communities. Such prediction usually relies on net-demand information, but each building can have additional information, such as separated generation and demand profiles, weather, or occupancy data. Such information is not only relevant for the net-demand prediction of each building, but also to other buildings with the same type of use. However, buildings avoid sharing such information due to privacy concerns. This paper proposes a novel federated learning framework for predicting building temporal net-demand in transactive energy communities. The proposed approach leverages centralized oversight of a central agent (aggregator) to inform distributed collaboration among each client (buildings), which are willing to collaborate to improve their prediction accuracy. The proposed approach was tested using a dataset collected from several buildings from a University campus (from the University of Coimbra in Portugal), predicting the electricity demand, and then using the local generation data to evaluate the net-demand, in the community of buildings.