P2P Energy Sharing with Federated Learning and Blockchain

Abstract

This paper presents a peer-to-peer electricity trading framework with dynamic pricing designed to enhance energy availability and economic returns for market participants. A key feature of the proposed model is the use of federated learning for predictive pricing for energy, ensuring data privacy while considering user preferences. Simulation results validate that this approach significantly improves energy availability and financial benefits. The method is implemented on the Ethereum public blockchain, providing a secure and transparent trading environment. Real-world data tests further confirm the efficacy of the system, demonstrating enhanced reliability and economic gains for participants in decentralized energy markets.