Efficient Assignment of Electric Vehicles to Charging Stations

In a large-scale integration of electric vehicles (EVs) to the transportation sector, it becomes more challenging to provide an acceptable level of service to EV users who already suffer from EV's short range and long charging time. Because of the latter, charging facilities can be easily overloaded, leading to intolerable service delays. Achieving an acceptable level of service will require an extensive upgrade for power networks. However, it is necessary to make sure that existing charging resources are efficiently utilized before taking such expensive decision. In this paper, we provide a user-oriented EV control scheme based on the efficient assignment of EVs to charging stations. The key performance indicator is the average time spent by the user from requesting the charging service to accessing it. Our direction is thus opposing to that of the majority of EV-related studies which sacrifice user convenience for achieving utility benefits. The methodology we use depends on defining a queueing model to facilitate the management of a dynamic population of EVs, then developing an EV assignment algorithm based on the Lyapunov optimization method. Our scheme is compared to that of other two dynamic assignment schemes which are the nearest-station policy and join-the-shortest queue policy.