About the project
The national grid is increasingly struggling to cope with the increasing renewable energy supply, as well as the increased demand from electric vehicles. To address this challenge, this project will use multi-agent system approaches including mechanism design and MARL to design decentralised systems whereby energy is produced and used locally.
The national grid is struggling to cope with the increasing renewable energy supply, as well as the increased demand from electric vehicles and electrification of households and businesses. As a result, many households and businesses have restricted production capacity and are facing long delays waiting for network upgrades. At the same time, new opportunities such as hydrogen are enabling easier and cheaper energy storage in the near future.
To address this challenge, this project will use multi-agent system approaches including incentive engineering (i.e., mechanism design) and multi-agent reinforcement learning (MARL) to design novel decentralised systems whereby energy is produced and used locally. In doing so, a major challenge is designing mechanisms which are both efficient and fair, meaning the matching of local demand and supply is done in a way that is equitable and satisfies important fairness properties such as envy-freeness, whilst ensuring local grid capacity and voltage constraints are met.
During your PhD, you will be part of an inclusive, vibrant and supportive research group. The agents, interaction and complexity (AIC) research group at Southampton University is one of the largest research groups in the UK focusing on multi-agent systems and AI, and you will be supervised by experts in energy and multi-agent systems.
[1] Alabdullatif, Abdullah M., Enrico Gerding, and Alvaro Perez-Diaz. "Market design and trading strategies for community energy markets with storage and renewable supply." Energies 13.4 (2020): 972.
[2] Buermann, Jan, Enrico Gerding, and Baharak Rastegari. "Fair Allocation of Resources with Uncertain Availability." AAMAS 2020.
