This chapter proposes a stochastic optimization-based energy regulation method for fast charging station aggregators participating in two-stage electricity markets, addressing the challenges of multidimensional uncertainties and bounded rationality in EV user behavior. A path merging and flow splitting strategy is first introduced to simplify the bounded rationality dynamic user equilibrium model, enabling efficient prediction of spatiotemporal charging load distributions in large-scale networks while significantly reducing computational complexity. A group-based charging scheduling strategy is then developed, clustering EVs with similar charging requirements to substantially decrease decision variables. An aggregator model incorporating local renewable energy and storage is formulated, with K-means clustering used to generate representative scenarios for travel demand, real-time prices, and renewable output uncertainties. The stochastic optimization framework determines optimal day-ahead bidding and real-time charging strategies. Case studies based on empirical data from a provincial electricity market in China demonstrate that the simplified model reduces iterations by 68.8% and computation time by 94.9% while maintaining high accuracy. The proposed strategies increase aggregator profits by up to 14.7% through peak load shifting, reduced deviation penalties, and enhanced renewable integration. The method exhibits excellent scalability for large-scale EV applications, providing theoretical and technical support for aggregator participation in electricity markets.