A plug and play artificial intelligent architecture for smart local energy systems integration

Euan Morris and Stephen McArthur

11th October 2021

For smart local energy systems (SLES) to be truly ‘smart’ they must be capable of managing complex interactions between data, users and the physical devices which make up the network, while also meeting the key criteria of flexibility, scalability, and reusability.

To investigate the potential for Plug and Play Artificial Intelligence techniques to help in meeting these aims, a Multi-Agent Systems (MAS) approach was taken to facilitate the trading of energy between a range of simulated SLES units. This solution was successfully deployed on top of the existing control system for each of these units.

This investigation showed the following:

• The use of MAS to allow inter-unit energy trading brought down the costs of balancing the individual smart local energy systems.

• The MAS solution was demonstrated to build scalability in the intelligent power management system.

• The MAS architecture design took a flexible PnP approach, which could accommodate market participants leaving and joining over time, and could accommodate a range of technology options.

• The MAS design provide for reusability, and could be deployed in a number of locations where SLES are interconnected.