Mixed game-based optimal scheduling for distribution network with multiple virtual power plants leasing shared energy storage

Large-scale renewable energy penetration forces the traditional distribution network to transform into the active distribution network (ADN). Collaborative sharing of energy storage and virtual power plant (VPP) under a sharing mode facilitates the coordinated economic operation on the supply side and enhances flexibility in grid-side adjustments. Therefore, a mixed game optimization scheduling method for ADN with multiple VPPs leasing shared energy storage (SES) is developed in this paper. Specifically, the ADN is modeled as the leader in a Stackelberg game optimization framework, while a cooperative alliance comprising multiple VPPs and SES operators (SESOs) performs as the follower. A bisection-based method is employed to address the Stackelberg game model, yielding the optimal time-of-use pricing and operational strategy of the alliance. Considering electricity interactions and cost-sharing among alliance members, a cooperative game model involving multiple VPPs and SESO is developed based on Nash negotiation theory and further divided into two sub-problems. To resolve sub-problems efficiently, a distributed algorithm leveraging the adaptive alternating direction method of multipliers (A-ADMM) is developed. Simulation results show that the peak-valley reduction effect of the ADN is significantly improved, the net load standard deviation is reduced by 9.23 kW, and the peak-valley ratio is reduced by 8.62%. Also, the operating costs of the three VPPs decreased by 3.67%, 11.44% and 7.00% respectively, and the operating income of SESO increased by ¥720.96. The validity and practicality of the proposed method are demonstrated.