Robust Optimization for Stackelberg Game of Multiple Virtual Power Plants Aggregated with Electric Vehicles

Abstract

In the new power system integrated with high-penetration distributed renewable energy and electric vehicle (EV) clusters, traditional scheduling methods cannot adequately address the game interactions among multiple virtual power plants (VPPs) and the stochastic fluctuations in EV operational behaviors. This paper constructs a bi-level optimization framework based on the Stackelberg game theory. The upper-level model optimizes transaction electricity prices to maximize the profit of the VPP operator, while the lower-level model realizes internal scheduling for each sub-VPP with the objective of minimizing operational costs. This study innovatively integrates aggregated EV resources into the multi-VPP game framework. It characterizes the operational features of EVs, including travel demands and charging-discharging behaviors, and adopts adjustable robust coefficients to achieve quantitative management and control of uncertain risks. Case study results indicate that the proposed strategy can fully exploit the peak shaving potential of EVs. The dynamic pricing mechanism facilitates transactions among VPPs and effectively reduces the operational costs of all participants. Moreover, the robust optimization method significantly enhances the system’s anti-disturbance capacity, which enables the system to adapt to complex practical operation scenarios.