Risk-averse two-stage battery scheduling for a hybrid charging network including centralised and on-site charging operations
研究了混合电动汽车电池交换网络，通过规避风险的两阶段随机规划模型优化集中充电和现场充电调度，在需求不确定下降低成本并提升服务。
Battery swapping has emerged as a fast and flexible alternative to plug-in charging for electric vehicles (EVs), especially in high-demand urban settings. However, operating an efficient swapping network poses infrastructure constraints, particularly when local charging capacity is limited. This paper investigates a hybrid EV battery swapping network that integrates centralised charging at a depot with on-site charging at local swapping stations. We develop a risk-averse two-stage stochastic programming (RTSP) model to optimise centralised charging, battery dispatch, and on-site charging under uncertain customer demand. To hedge against distributional ambiguity, we employ an L1-norm based uncertainty set and solve the model using a stabilised adaptive multi-cut L-shaped algorithm. We conduct a case study leveraging real-world battery swapping station (BSS) data from an existing network to illustrate the operational details with a modelled central charging station (CCS). Compared with the prevailing on-site-only strategy, the hybrid strategy delivers notable cost savings and service improvements, particularly when battery inventory is high, local charging capacity is limited, and centralised charging electricity prices are low.