Based On Ride-sourcing System Of Vehicle Relocation Problem Research

With the continuous popularization of the mobile Internet,concepts such as "sharing economy" and "low-carbon travel" have been applied in the form of Internet products,and gradually integrated into public life,such as shared rental Airbnb,shared travel Uber,etc.Among them,the most representative application field is transportation.Ride-sharing can not only meet the daily travel needs of some people,but also shared modes such as carpooling and ride-hailing can alleviate traffic pressure and reduce resource waste.In addition,this model has been continuously accepted and recognized by the public and has become one of the indispensable modes of transportation.However,the ride-sharing model has an imbalance between vehicles and demand in actual operations.If this problem can be dealt with,it will not only bring a better customer experience,reduce costs and increase efficiency,but also help relieve traffic congestion.Therefore,this paper proposes the vehicle relocation algorithm system framework in the ride-sourcing mode,designs three vehicle relocation strategies and strategy evaluation indexes,and evaluates the three strategies under different scenes and vehicle inputs.Specifically,the three relocation strategies and algorithms are: the first is the Euclidean distance relocation strategy,which is an integer programming model that maximizes the trip request as an objective function.The core is based on the distance between the vehicle and the request.The second is the trip request rate relocation strategy,that is based on the dispatch of empty cars generated in the current period to the waiting request through design trip request rate;the third is the dynamic vehicle relocation strategy,namely based on the execution of the trip request rate relocation strategy,each remaining empty car is decided to dynamically idling or waiting,and the cars in the idling process can participate in matching.In addition,this article studies three scenarios and algorithms.First,the timespace uniform scene refers to the uniform distribution of any time period and any place within the system;second,the time uneven scene refers to the fact that the number of trip requests in certain time periods is significantly higher or lower than the rest of the time segment;Third,the spatially uneven scenes refer to certain areas where the number of trip requests is significantly higher or lower than other areas.By designing passenger travel time,empty waiting time,vehicle relocation time,total consumption time,passenger travel distance,vehicle relocation distance,total travel distance,ride request(passenger)waiting time,ride request service number and idling time indicators to evaluate the three strategies.The system operation is simulated through simulation,and the index results presented after different relocation strategies are adopted in different scenarios,which provides a basis for vehicle relocation decisions and relocation strategy selection.