Two-stage stochastic sizing and packetized energy scheduling of BEV charging stations with quality of service constraints

The expected deployment of battery electric vehicles (BEVs) strongly depends on the development of an adequate charging station (CS) infrastructure that guarantees a certain level of quality of service (QoS) to the BEV users. This paper proposes a two-stage method to select the number and type of CSs in parking areas (PAs) and schedule the charging sessions of the incoming BEVs ensuring a predetermined QoS level while minimizing the cost for the CS manager. In particular, stage one solves the CS sizing problem while stage two involves a probabilistic simulation procedure able to solve the charging scheduling problem by using a packetized energy approach. We also take into account the typical charging current and voltage characteristic of a BEV battery pack, and the real statistical distribution of BEV arriving times and expected parking times. A case study based on the PA at the University of Salerno Campus is used to demonstrate the effectiveness of the proposed method.