Preserving fairness in EV charging under time-varying congestion levels

Abstract

This paper concerns facilities for charging Electrical Vehicles at parking lots. We work under the assumption that power capacity may be insufficient to simultaneously charge all stations, and thus some scheduling must be performed. In previous work we analyzed the case of a stationary customer demand, developing a fluid model which characterizes the performance of different scheduling policies in overload. A new policy termed Least Laxity Ratio was proposed to improve fairness in service. In this paper we wish to incorporate the fact that congestion levels are not stationary in practice, rather they obey daily use cycles. We study the behavior of the different policies with load obtained from a real set of parking lot data. Empirical results show that the conclusions of the stationary analysis remain valid. In particular, during intervals of congestion, LLR achieves the best results in terms of proportional fairness.