摘要：AbstractElectrical Vehicles (EVs) have demonstrated significant fuel saving benefits over conventional vehicles. DC fast charging stations (DCFCs) have become the dominant charging option when fast charging speed is needed or residential charging station is unavailable. The cost of charging EVs using DCFCs may vary significantly due to the strong nonlinearity of the charging power and a relatively higher charging cost than the domestic charging scenario. Therefore, optimal charging schedules that lead to minimal charging cost, are of great interests to EV users with dramatically explosive EV adoptions in the market. In this paper, two global optimization algorithms, Genetic Algorithm (GA)-based and Dynamic Programming (DP)-based, are proposed to optimize the EV charging schedule at DC fast charging stations to minimize the charging cost, provided that day-to-day EV data in real-world operation are predictable. Compared to the non-optimal charging schedule, GA-based and DP-based optimal charging schedules can reduce the charging cost by 42.7% and 46.3%, respectively, in a 13-day application. With the proposed optimization algorithms, both the EV charging time and final state of charge (SOC) at the end of charging in individual charging events can be globally optimized to minimize the charging cost.
