Multi-agent systems in Peer-to-Peer energy trading: A comprehensive survey

Energy networks around the world have experienced a significant increase in the amount of distributed generation. This decentralization of energy markets has led to a surge of interest in Peer-to-Peer energy trading between individuals. There are many challenges associated with Peer-to-Peer energy trading, e.g. behaviour modelling, decision making and optimization. Multi-agent systems is a prominent subfield of Artificial Intelligence research that is very effective for these types of problems. This research aims to provide a detailed survey of recent advances in the application of multi-agent systems in Peer-to-Peer energy trading. The challenges encountered in implementing these systems are discussed, e.g. agent learning, privacy, and computational power. The primary advantages of multi-agent systems for Peer-to-Peer energy trading reported in the literature are: improved efficiency of renewable utilization, reduced transaction costs, and scalability. Current challenges and future directions of research are also outlined.