Multi‐leader–follower game theory for modelling interaction between virtual power plants and distribution company

Virtual power plant (VPP) is an entity that aggregates generation of distributed generation units. Thus, it is important to design a competitive framework which models the participation of the VPPs in the electricity market along with their trading with distribution company (DisCo). This study proposes a bilevel programming framework using the concept of multi-leader-follower game theory to determine the optimal contract prices of VPPs which compete against each other in the distribution network. The optimal prices are used for setting annual bilateral contracts with VPPs. The leader layer of the proposed bilevel problem includes VPPs, which try to maximise their profits, while the follower problem corresponds to the cost function of the DisCo, which aims to minimise the payments of supplying the forecasted demand. The DisCo optimisation problem is transferred by its Karush-Kuhn-Tucker optimality conditions, turning each VPP problem into an equivalent single-level optimisation problem. Some case studies are defined and implemented to assess the performance of the proposed scheduling model.