Planning Stand-Alone Electric Vehicle Charging Stations on Highways

This chapter puts forward a comprehensive two-stage method for determining the locations and capacities of stand-alone electric vehicle charging stations on highways. In the first stage, we determine the locations where individual vehicles necessitate charging services via a Monte Carlo simulation, which utilizes traffic demand and battery data. Subsequently, we propose an integer programming model to pinpoint the optimal locations for charging stations to guarantee that every vehicle has access to at least one charging station without depleting its battery. In the second stage, a distributionally robust optimization (DRO) model based on Kullback-Leibler divergence is constructed to optimize the capacities of renewable generation and energy storage units within each charging station. Two reformulations of the DRO model are proposed. The first reformulation, based on value-at-risk, results in a more accurate mixed-integer linear programming problem. The second reformulation, based on conditional value-at-risk, provides a more tractable, albeit conservative, linear programming approximation. This chapter introduces the research presented in Xie et al. (2018).