Electric vehicle (EV) charging networks serve as a critical interface between transportation electrification and power systems. Configuring and operating charging infrastructure has become a key decision problem for EV charging network systems. This paper investigates the integrated capacity configuration and operational scheduling of an electric vehicle charging network that incorporates conventional charging service and vehicle-to-grid (V2G) interaction. The objective is to balance long-term capacity investment with short-term operational decisions while accounting for EV users' charging and discharging behaviors and renewable power output uncertainty. A two-stage stochastic programming model is proposed to jointly determine charging station capacity, multi-source power allocation, and EV charging and discharging schedules. Uncertainty is represented by scenarios generated using Latin hypercube sampling, and the resulting large-scale stochastic program is addressed through a sample average approximation method combined with the progressive hedging algorithm. Numerical experiments based on real-world data from Beijing show that V2G operations significantly enhance operational flexibility, increasing peak shaving and valley filling capabilities by approximately 151.8% and 43.6%, respectively. Sensitivity analysis reveals that charging and discharging prices are crucial for regulating user participation, while higher station-level energy self-sufficiency reduces dependence on strong price incentives. Furthermore, coordinated integration of energy storage with renewable supply improves renewable energy utilization and alleviates the temporal mismatch between energy availability and charging demand. These results offer theoretical support for the configuration and scheduling of urban electric vehicle charging networks, thereby enabling high renewable energy penetration and supporting power system decarbonization.
Yue Wang (Fri,) studied this question.