Optimal planning of charging stations for an electric vehicle fleet in car-sharing business

With the increasing market share of electric vehicles (EVs), optimizing the planning decisions of charging stations becomes critical to support long-distance traveling. Besides, the booming car-sharing industry provides an alternative business model to enhance the utilization of vehicle resources. Recently, several car-sharing enterprises have launched the collaborations with charging station construction companies to construct their own charging facilities, which calls for the co-optimization of charging station planning and scheduling of EVs in car-sharing business. Therefore, this paper develops an optimization model for electric vehicle charging stations placements and type selections, considering routing selection and charging management of an electric vehicle fleet in car-sharing business. Specifically, a stochastic programming framework is exploited to incorporate a variety of operating scenarios, which reflect the uncertainties of riders' requirements, drivers' itineraries, traffic flow information, and electricity prices. The proposed model aims to improve the welfare of all riders and reduce total cost of fulfilling riders' pickup and delivery requirements simultaneously, in which determining the optimal locations and types of charging facilities is required. Moreover, portfolio management is integrated into the optimization model to mitigate investment risk. After a series of reformulation, the proposed optimization model can be recast into a mixed-integer linear programming problem. Using two transportation networks, the effectiveness of the proposed model is validated in numerical studies. It finds that the optimal electric vehicle charging station planning strategy is determined according to the operator's preferences, electricity prices, and installation budgets. Moreover, diversity within operating scenarios yields the positive value-of-stochastic-solution that demonstrates the necessity of implementing stochastic programming.