Adaptive price control for electric vehicle charging in smart grid

Smart grid, the future of power grid, aims at making progress on electricity reliability and emission reduction. In recent years, electric vehicles (EVs) are adopted increasingly due to their zero discharge and high efficiency. For the reliability of smart grid, the charging control with high penetration of EVs is needed to prevent from overload and power loss. In this paper, the adaptive price control is proposed for EV charging. The aggregator manages the EV batteries and regulates the charging demand with the consideration of energy supply limit by using price control. We consider that the information of EV mobility is unknown in advance, which will impact the performance of price control. Thus, the technique of Neuro-dynamic Programming (NDP) is leveraged to obtain optimal price policy by processing online learning. Numerical results show that our adaptive price control can tune the EV charging demand to approach the expected level by learning from the EV charging process and the EV mobility.