Abstract:

In recent years, the growing penetration of electric plug-in vehicles has called for specific solutions relating to charging policies to be used in parking lots equipped with charging stations. In reality, simple policies focused on uncoordinated vehicle charging usually lead to a high peak demand for electricity, which can lead to high costs for the owner of the car park. The aim of this work is to design a charging power schedule for each unit able to minimize the overall daily peak power while satisfying the customer requirements in terms of charged energy. Three sources of uncertainty affect the incoming vehicles: the arrival time, the departure time and the demanded energy to be charged. To assess the quality of the charging service under such uncertainties, a suitable customer satisfaction policy is employed. Depending on the information available on the uncertain variables, two algorithms based on a receding horizon approach are designed. Such algorithms require the solution of linear programs and provide the charging power for each plugged-in vehicle. To evaluate the performance and the computational feasibility of the proposed techniques, numerical simulations have been performed and comparisons are provided versus the uncoordinated charging policy and an ideal (omniscient) algorithm able to provide a lower bound on the daily peak power.