Location Of Charging Station Under Risk Aversion

In recent years,China’s new energy vehicle(EV)industry is in a period of rapid development,and the annual sales of new energy vehicles from 2016 to 2022 ranked first in the world for six consecutive years.With the rapid growth of the number of electric vehicles,the number of charging stations,as supporting infrastructure for new energy vehicles,also shows a rapid growth trend.In 2020,the number of charging stations in China has reached 4.8 million.The construction of infrastructure is a process with a long construction period and cost of money.It is very important for charging station facilities to meet the demand as far as possible when facing the charging demand of rapid expansion.Therefore,it is necessary to make reasonable planning of charging station facilities by scientific means.In order to further study the facility location problem of charging stations,improve the rationality of charging station site selection,and fully consider the uncertainty of charging demand,this paper takes electric vehicles as the research object,solves the random distribution of charging demand by establishing a multiscenario approach,and establishes a two-stage stochastic programming model with the goal of maximizing the charging demand that can be met by the combination of charging stations.Factors such as route choice behavior of EV users and vehicle driving distance are integrated into the decision-making process of the model.In the model based on user shortest path behavior,the first stage takes the maximum charging demand covered by charging station facilities as the goal,and the upper limit of the number of charging station facilities as the constraint condition.In the second stage,the concept of conditional risk value is introduced,aiming at maximizing the average coverage demand of multiple worst-case scenarios,the model is constructed with the corresponding relationship between road arc and charging station as the constraint condition.In the model based on the user’s deviated path behavior,the degree of the user’s deviated path and the mileage of the vehicle itself are considered,and the model based on the shortest path behavior is extended,so that the model itself is closer to the charging problems encountered by the user in the real journey.For the stochastic programming model constructed in this paper,sample average approximation method and two-stage method are used to solve the problem.Then,sensitivity test and numerical analysis are used to show the influence of various factors on the model results.Finally,in order to verify the effectiveness of the proposed model,the actual road network in Wenjiang District of Chengdu City is tested as an example.The results show that if a limited number of charging stations are to be built,the model in this paper is significantly better than the deterministic model in some important indicators of risk assessment,which means that compared with the deterministic model,the model in this paper will meet more EV charging demands.From the perspective of management,considering risks at the initial stage of building charging stations is conducive to dealing with the uncertainty of future charging demands,thus ensuring the long-term and stable operation of charging stations.