Congestion-aware deployment of electric vehicle charging stations

Nowadays, Electric Vehicles (EVs) have gone through a rapid development since people pay more attention to environmental conservation. However, the availability of EVs is seriously restricted by the scarcity of efficient charging infrastructures. In this paper, we propose a congestion-aware charging scheme to deploy Charging Stations (CSs). We first leverage vehicles' dwell events to define the charging demand. Concretely, we take advantage of M/M/x/N queueing model to calculate the probability of a vehicle being rejected by a CS. We further formulate a charging rejection probability minimization problem based on assessable information of CSs. We have proposed heuristic algorithms to determine the optimal deployment strategy. Finally, we perform extensive experiments using a real taxi trajectory dataset. Results demonstrate that our proposed algorithms are effective in minimizing vehicles' charging rejection probability compared with baselines.