Game Theoretic Orchestration for Cooperation among Power Distribution System Applications

The ongoing transformation of the grid necessitates advanced distribution systems to integrate and orchestrate a large number of grid-edge devices while also serving multiple system-level objectives such as resilience, decarbonization, equity and other system mandates. The parallel deployment and control of resources towards achieving diverse objectives may lead to conflicts between applications that want to control overlapping sets of device setpoints. Unaddressed, these conflicts can lead to oscillatory behavior and sub-optimal performance. This work introduces a game theoretic framework to drive cooperative behavior among competitive applications within a modular platform. The work proposes a weighted-consensus-based game design to facilitate conflict resolution through consensus-building iterations. Simulation-based evaluation demonstrates that the proposed deconfliction strategy can resolve operational conflicts and achieve close-to-optimal trade off among the applications. Results also compare the proposed strategy with a distributed optimization approach that illustrates the effectiveness of the proposed approach. The approach demonstrates that a scenario involving participating and non-participating apps resulted in a 33.5% improvement in worst-case normalized app-objective compared to the distributed optimization approach.