Emerging AI Technologies for Decentralised Energy Systems

Abstract :

Energy systems are becoming increasingly decentralised, with many autonomous parties (e.g. prosumers, small generators, distributed battery systems etc), while at the same time required to deal with new challenges, such as intermittent renewable generation, and new forms of demand – for example, from large-scale electric vehicle charging. The talk will give a broad overview of how multi-agent system models and artificial intelligence techniques can model and address these challenges. The talk will cover applications, ranging from the charging of self-interested electric vehicles, such that the charging capacity of local distribution networks is not exceeded, or routing EVs at distributed charging points to minimize queuing times. The talk will also briefly touch on some other scenarios where incentive design plays a key role, such as distributed demand-side response, virtual power plants, or peer-peer trading and fair sharing of resources in energy communities.