P2P Energy Trading for Interconnected Microgrids: Genetic Algorithm with Hawk-Dove Battery Usage Strategies

Peer-to-peer energy trading among interconnected microgrids can improve local renewable energy utilization and reduce dependence on the main grid. However, existing approaches do not fully integrate participant specific battery use behavior, degradation aware pricing, energy allocation, unmatched energy reduction, and AC network feasibility. This paper proposes a genetic algorithm-based P2P trading framework for interconnected microgrids equipped with battery energy storage systems. At each trading interval, microgrids are classified as sellers, buyers, or neutral participants and assigned aggressive or conservative battery use strategies inspired by the Hawk-Dove game. Participant specific marginal energy values are derived from Lagrange multipliers and adjusted according to battery strategy and degradation cost, while a genetic algorithm determines bilaterally exchanged energy quantities by minimizing trading cost, battery degradation cost, and unmatched energy. Candidate solutions are re-paired to satisfy battery operating limits and bus-voltage and branch-loading constraints. The framework is evaluated using 5-minute renewable generation and load data for 20 microgrids connected through an IEEE 33-bus system. The results show that matched energy exceeds unmatched energy during most trading intervals, including under demand dominated conditions. Most bilateral prices provide economic benefits to both sellers and buyers, resulting in positive network level welfare on two representative days. The retained trading schedules also satisfy the considered battery and AC network constraints.