The Location-sizing Problem Of Electric Vehicle Charging Station Deployment Based On Queuing Theory

With the ever increasing problem of environmental pollution and the oil crisis, the electric vehicle (EV) with saving energy and decreasing emissions has become more and more popular. With the EV technology’s improvement and the government’s support, the EV market grows constantly. Meanwhile, the imperfect EV charging infrastructure network inhibits the promotion of the EV. So, building a good charging infrastructure network is very important for promoting the use of electric vehicles. Compared with traditional fuel vehicles, electric vehicles have two obvious disadvantages:short travel range and long charging time, so how to decide the location and sizing of the charging station is very important. For charging facilities network construction, the location of sites and the number of equipped charging spots in charging station directly affects the quality of charging service and operating efficiency. Given a specific budget fee, if we establish more charging stations, then the charging spots equipped in charging station will be less. That is, the coverage level increasing at the sacrifice of increasing waiting time. If we allocate more charging spots in one charging station, then the number of charging station will decrease. So, the waiting time decrease with the lower coverage level. Therefore, how to determine the location of charging station and the sizing of the charging spots is very important.Based on the location problem, we can divide the charging station deployment into two types:charging station deployment on the highway network and the charging station deployment on the metropolis. According to the location studies, we can divide the location-sizing problem into two types based on the charging demand:path demand and node demand. Combined with the electric vehicle charging station queuing theory model, we developed different location-allocation model for electric charging station under different demand patterns. For the path-based demand patterns, we formulate a location-allocation model maximizing the number of EV that can be recharged in the path, considering the upper limit of waiting based on the knowledge of flow-refueling location problem and the queuing theory. For the charging demand caused in nodes, expectancy model was applied to study the EV charging station location-allocation problem. Given the max response time, we formulate a model to maximize the expected charging demand. Then, we did some sensitivity analysis with numerical experiments to analyze the impact with result of different impact factors.