Inventory management of battery swapping and charging stations considering uncertainty

Abstract

The battery swapping mode of electric vehicles (EVs) is expected to play an essential role in transportation and power systems. Plenty of batteries are managed by the battery swapping and charging stations (BSCSs) as the core assets. Proper inventory scale and charging/discharging control of batteries are essential to the planning and operation of BSCSs. This paper addresses two critical aspects of BSCS batteries: the optimal amount of inventory batteries and the tradeoff between the initial number of depleted and fully-charged batteries at the beginning of each scheduling horizon. Aiming at these, the time coupling between initial inventory, charging/discharging strategy, and swapping time sequence is analyzed. Afterward, it establishes the initial inventory based charging/discharging (IIC) model. The uncertain swapping demand is fully considered by utilizing a data-driven distributionally robust optimization (DRO) approach with an improved K-means clustering algorithm. A duality-free column-and-constraint generation (C&CG) algorithm solves the problem iteratively. The simulation results demonstrate the validity of the IIC model and show that the obtained optimization results achieve the balance of economy and robustness. Moreover, the impact of battery technology development on inventory batteries is analyzed, and suggestions are given.