Optimized Adjustment Strategy Of One-way Car-sharing System Based On The Characteristic Mining Of User Demand

Car-sharing has become a new choice with the rapid development of the Internet.The operation mode of car-sharing can be divided into station-based and free floating,in which the station-based car-sharing system can be divided into one-way and round-trip.The round-trip car-sharing system is a traditional car rental model,which requires users to return the car at the same station after borrowing the car at the station designated by the operator.One-way car-sharing system means that users can return the car at any station in the system after borrowing the car at the station designated by the operator.Free floating allow users to borrow and return cars at any station,including public parking spaces.Considering the convenience of users and the feasibility of operation management,the one-way car-sharing system with fixed stations is the best way to take both sides into account.Based on this,this paper studies the optimization and adjustment of station layout for one-way car-sharing system on the basis of deeply mining the characteristics of user requirements under the current operating state.First,through the analysis and study of the operation data of the car-sharing system,combined with the current station layout scheme,the problems of low utilization ratio of vehicles and parking spaces and low revenue of the system are found.By analyzing the situation of borrowing and returning cars at different stations,we find that the supply and demand of the stations are not balanced.Second,the trajectory data of vehicles is used to mine the potential demand for shared vehicles.Combined with the random demand generated based on the analysis results of the borrow and return vehicles in different functional areas,the obtained final demand data is taken as the parameters of the model.And then,considering the current station layout of the system,a mixed integer linear programming model is established to determine the optimal layout adjustment scheme and the actual requirements of the station-based one-way car-sharing system,the algorithm is solved by Cplex.The results show that 84% of the original stations waste resources and need to reduce the scale.Third,a bi-level model is proposed to solve the station selection problem of new entrants in the case of operators already existed in the car-sharing market.The model fully considers a variety of factors that affect the user’s choice,and uses the logit model to describe the user’s choice behavior.In the bi-level model,the lower-level model is a convex optimization problem with linear constraints.In this paper,it is transformed into a single-layer model by KKT condition.Finally,the commercial solver Cplex is used to solve the problem,and the station layout plan of the newly added operator and the satisfaction of the requirements of the two operators are obtained.The solution results show that newly joined operators should adopt low-price strategies when entering the market to attract more users and realize revenue.