Day-ahead Multistage Stochastic Optimization of a Group of Electric Vehicle Charging Stations

This paper deals with the day-ahead optimal scheduling of a parking lot with several bidirectional charging stations for plug-in electric vehicles (EVs), part of a grid-connected system including also a photovoltaic (PV) generating unit and local loads. In the proposed approach, a central dispatching unit implements a multistage stochastic optimization to obtain the day-ahead scheduling of the charging stations considering the uncertainties associated with PV generation, non-dispatchable loads and the connected electric vehicles. The scenario tree is built by means of a reduction technique based on k-medoids so that all the representative scenarios included in the tree are feasible. The objective is the minimization of the expected daily procurement costs, exploiting the Vehicle-to-Grid (V2G) services provided by the parking lot. The performance of the procedure is assessed by using different case studies.