Optimization Model And Algorithm Of Car-sharing Station Location

As a new type of transportation,car-sharing has been rapidly developed with the development and support of the Internet and the Internet of Things,and can meet the diverse travel demands of users.At present,there are three main types of car-sharing mode:one-way,round-way and free-floating.The one-way system allows vehicles to be returned at any station,the two-way system requires users to return the vehicle to the station where the car is rented,and the free-floating system allows users to return the vehicle at any place where parking is allowed.These characteristics determine that the free-floating type is the most flexible mode.However,since car-sharing have not been popularized,the one-way system is the most commonly used model at present.Based on this,this paper studies the problem of location selection in the one-way car-sharing system.In this paper,on the basis of rent data and vehicle trajectory data,in-depth analysis of car-sharing stations and demand characteristics,on this basis,two car-sharing site location models are proposed,and case analysis is carried out.First,based on the analysis of data such as stations and demands,a data-driven location selection model was constructed.This model determined the degree of satisfaction of the demands through the distribution of potential demands,and took into account the degree of vehicle utilization,Finally,a mixed integer programming model is established.Solved by Cplex,the results show that the demand satisfaction rate is the most important factor in location selection,and vehicle utilization rate largely determines the number of parking spaces and vehicles.Secondly,in view of the fact that car-sharing companies generally choose electric vehicles,a location model considering charging time is constructed.Since the model takes into account the charging time and involves changes in time,a dynamic location model is designed.Aiming at the proposed model and combining the characteristics of the model,an improved genetic algorithm is proposed.This algorithm improves the basic genetic algorithm from the aspects of fitness function,crossover and mutation operations,and improves the solution efficiency.Taking Beijing as an example,the model is used to optimize the location selection.The results show that the charging efficiency(electricity consumption rate/charging rate)of electric vehicles has a greater impact on parking spaces,vehicles and energy consumption.And it can be adjusted according to the actual requirements.