Defence Xiupeng Chen: "Game Theoretic Coordination and Control of Decentralized Energy Systems"

Promotors: Prof. Nima Monshizadeh and Prof. Jacequlien Scherpen

Abstract: This thesis investigates game-theoretic coordination and control methods for decentralized energy systems, accounting for the self-interested behavior of independent entities while adhering to physical constraints and addressing information-sharing limitations. Firstly, it develops innovative energy management schemes to facilitate the procurement of flexibility services from Distributed Energy Resources (DERs) and constructs game-theoretic models to analyze the interaction dynamics of independent, self-interested market participants. Secondly, the thesis develops fully distributed Nash Equilibrium (NE)-seeking algorithms for the formulated aggregative and generalized aggregative games. These algorithms eliminate the need for a central coordinator, allowing market participants to avoid sharing their true decisions. Instead, they only exchange estimates of aggregate quantities and auxiliary variables with their immediate neighbors. In the case of incorporating complex AC power flow constraints, it proposes a semi-decentralized algorithm, where participants only share non-private information with others. Finally, the thesis designs a data-driven dynamic intervention mechanism in network games, where the regulator directly formulates the incentive design policy based on historically observed agent actions and interventions, bypassing the need for intermediate steps such as utility function or network identification. In summary, by utilizing methodologies based on game theory, operator theory, and data-driven control, this research contributes to the development of secure and private control of decentralized energy systems and enhances their efficiency, reliability, and sustainability.

Dissertation