Bounds for optimal control of a regional plug-in electric vehicle charging station system

In order to support the increasing penetration of plug-in electric vehicle (PEV) users, a novel regional PEV charging station system with DC level 3 fast charging is proposed in this paper. To promote sustainable energy, the proposed system is designed to be equipped with a distributed energy storage system charged by wind generation, solar PV generation, and electricity from the power grid, which can simultaneously charge multiple PEVs. The objective of the proposed system is to minimize operational cost. Wind/solar PV generation and electricity market price are input state variables in this problem and are predicted by support vector regression (SVR). The uncertainties of the SVR models are analyzed using a Martingale Model Forecast Evolution (MMFE). Finally, bounds of the optimal operational cost in this problem are evaluated with two stochastic measures, which can be solved using the expected value problem and the wait-and-see solution. Bounds from experiments simulating models of the Dallas-Fort Worth metroplex show that the largest uncertainty in the system occurs during weekdays in the summer.