Research On Carpool Scheduling Method Based On Taxi Operation Data

Urban taxi is one of the essential means of transportation for public travel.Taxi is more and more convenient,fast and comfortable to meet the personalized travel needs of passengers.With the continuous increase of urban public transport demand,more public resources are needed to meet the public travel demand.As a travel mode that can effectively relieve traffic pressure,taxi sharing is favored by the public in recent years.It can not only effectively improve the use efficiency of taxis,but also reduce the operating cost of taxis,save taxi costs for passengers,reduce vehicle exhaust emissions,and improve the urban environment.Firstly,based on the analysis of the forms and classification of taxi pooling and the solution of taxi pooling scheduling algorithm,this paper establishes a taxi pooling scheduling model with the overall revenue of taxi and passengers as the objective function.The model considers the time windows,capacity and order of passengers and drivers,and takes into account the interests of passengers and drivers.In order to depict the taxi operation process,the vehicle driving process is divided into two states: carpool and non carpool.Carpool matching algorithm and taxi scheduling optimization algorithm are designed to solve the model.Carpool matching algorithm is the process of matching passenger request information with vehicle information and establishing carpool information set.Secondly,according to the taxi operation data of Beijing,the urban hot spots are counted,and the congestion degree of urban roads is distinguished.The ant colony optimization algorithm is used to find the carpool information at the no-load time for the trip assignment,and the taxi transfer probability is optimized within the constraints and the scope of the model.Finally,based on the GPS operation data of Beijing taxi,the test set composed of 67000 taxi information and 100000 taxi request information is tested by using python software,and the effectiveness of the carpool scheduling model and optimization algorithm is verified.Compared with the non carpool state,the key parameters affecting carpool are analyzed by using the control variable method.The results show that the model can effectively describe the taxi sharing process,and the ant colony optimization algorithm can effectively solve the problem of taxi sharing matching and scheduling.The simulation test results show that in the sharing state,the driver’s income increases by 5.28%,the passenger’s taxi cost decreases by 3.65%,the operating cost decreases by 5.46%,and the average number of passengers is 2.36.The success rate of carpool matching is less affected by the degree of road congestion and vehicle size,but the vehicle delay time and passenger waiting time have a greater impact on the success rate of carpool matching.The research results of this paper can be used in the research of urban taxi sharing scheduling problem,which can provide reference and suggestions for vehicle dynamic scheduling,the construction of online taxi sharing platform and taxi sharing.At the same time,it has certain positive significance for improving taxi operation efficiency,improving public travel willingness,easing traffic pressure and reducing urban pollution.