On the benefits of using self-organising Multi-Agent architectures in network management

As power systems grow they are incorporating ever-increasing numbers of distributed generation such as renewables but also various other flexible agents e.g. electric vehicles and energy storage, that can act both as loads and generators with the ability to provide a number of ancillary services. This evolution signals a migration of monitoring and control responsibilities often involving embedded intelligence. This work aims to expand on previous investigations considering frequency control and examines distributed communication and control architectures through the medium of Multi-Agent Systems (MAS) focusing on voltage control in a radial microgrid. The investigation assesses control and communication performance across a range of agent architectures against four selected performance criteria, and an increasing agent population. The objective of the paper is to demonstrate that network conditions can be used to determine the most appropriate control and communication architecture. Obtained results demonstrate the effectiveness that a dynamic, self-organising control and communication architecture can have on modern and future distribution grids employing a high number of distributed agents.