Optimal aggregation of a virtual power plant based on a distribution-level market with the participation of bounded rational agents

Virtual power plants (VPPs) offer system flexibility by aggregating various distributed energy resources (DERs) and simultaneously create profit opportunities for these DERs. A fair and scientific operational model serves as a crucial guarantee for promoting the economically efficient operation of VPPs. Given that DERs are typically managed by various agents, this study introduces a multi‑leader single-follower Nash-Stackelberg game model. This model is designed to facilitate the optimal aggregation of DER agents in a competitive distribution-level market involving multiple players. In the optimization modeling, the DER agents function as leaders, formulating their bidding strategies to maximize profits. Conversely, the VPP acts as a follower, responsible for performing market clearing. The distributed locational marginal price derived from this process serves as the settlement basis for DERs. Furthermore, a level-k reasoning approach is used to simulate the strategic bidding behavior of agents. Case studies demonstrate the effectiveness of the proposed aggregation model in simulating the competitive dynamics among agents. This model offers an expanded set of Nash equilibrium solutions for VPPs, enabling the selection of suitable aggregation schemes tailored to practical requirements. Additionally, the proposed bounded rationality model mitigats the strategic bidding behavior of agents, which consequently leads to a reduction in the system cost.