FLEDGED

The FLEDGED project aims to optimize the integration of renewable energy into low-voltage networks using federated machine learning. By employing federated learning techniques, local data from decentralized systems such as PV installations, heat pumps, and battery storage can be analyzed. What makes this approach unique is that the data processing takes place directly at the individual device level, avoiding the transmission of large data volumes and thus addressing data privacy concerns.
The project’s objectives include the development, testing, and evaluation of federated learning techniques for specific low-voltage applications in distribution systems, such as the enhancement of services for flexibility forecasting, supporting optimal energy market decisions, avoiding costly grid reinforcements, and improving grid stability.