Demand side energy management of EV charging stations by approximate dynamic programming

The high operation cost of the EV charging station (EVCS) is a severe challenge for the development of electric vehicles, which lead to the general shortage of the EVCS. In order to reduce the operation costs of the EVCS, an approximate dynamic programming (ADP) based energy management system (EMS) is proposed for the EVCS equipped with multiple types of chargers (EVCS-MTC). A fuzzy logic guiding system has been designed to allocate each vehicle an appropriate charging spot based on its charging urgency. Multiple EVs can acquire the charging service through a common charger in the EVCS-MTC. In the proposed EMS, the approximate dynamic programming (ADP) and the evolution algorithm (EA) are combined to determine the optimal charging start time for each EV. This characteristic provides the charging device with the maximum autonomy to select the preferred flexible charging pattern, which can prolong the battery lifetime and reduce the communication requirements of the control system. With taking the dynamic electricity price and uncertain future charging demand into account, the proposed EMS can achieve a total cost reduction of over 50% compared with the conventional charging scheme in the numerical studies.