Day-Ahead Planning for EV Aggregators Based on Statistical Analysis of Road Traffic Data in Japan

Electric vehicle (EV) is expected to penetrate all over the world in the future, which provides EV aggregators with business opportunities. They can produce profits by buying and selling electricity from/to the electric power market through vehicle-to-grid (V2G). It is vital for EV aggregators to deal with the uncertainties of EV such as its plug-in/plug-out time and necessary charging amount. Although many studies have been applying stochastic programming to tackle with the uncertainties, their model and scenarios presenting the uncertainties are not enough for the real problem. In this paper, the scenarios used for the problem are built accurately based on analysis of an enormous amount of survey data in Japan. Then the solution of stochastic programming is evaluated. This paper also refers to performance of the scenario reduction by backward scenario reduction technique.