Study On Passenger Classification And Timetable Optimization Of County Bus Based On Big Data

With the rapid development of China’ s new-type urbanization,a large number of people are flooding into the urban areas of which the scale has been expanding,resulting in a yearon-year growth of passenger traffic demand.Despite the fact that public transportation of the county owns has been improved notably in recent years,there are some passenger transportation lines where passengers are less and mainly commuters.The seating ratio is low for the most of the time and the bus companies suffer serious losses.However,expanding the interval between buses keeps passengers waiting longer,which has become a key point for these bus routes to continue to provide bus services.Based on the big data background of public transportation system,this paper conducts in-depth mining analysis of public transportation data,classifies passengers according to different classification criteria,and constructs a single-line bus schedule optimization model based on dynamic demand.Firstly,based on the existing research results of scholars at home and abroad,this paper deeply analyzes the problems existing in the current county bus operation in China.Relying on the public transportation data of Weichang County,we can forecast the passengers’ boarding station with data fusion methods.The passengers’ drop-off station can be deducted out with the probability model of the passengers’ off-site based on the travel chain;In addition,the time-constrained method is used to identify the transfer passengers and to get their OD data,then we define the data with the cumulative passenger flow demand function.Secondly,with the weekday’s IC card data,the travel time features of passengers with different IC card types can be analyzed from the overall and individuals respectively.According to the passengers’ travel intensity,the k-means clustering algorithm is used to realize the classification research of the passengers using credit card and to dig out the proportion of different passengers.Combined with the questionnaire data,a SVM model is constructed to realize the identification of commuter passengers.Finally,by analyzing how the features of the parameters of the cumulative passenger flow demand function change with different departure timetables,a time probability selection model for passengers arriving at the bus station is established;and the parameters in the model can be selected out using the OD history data to obtain the cumulated passenger flow demand function.The calculation method of the bus departure schedule realizes the prediction of dynamic passenger flow OD;on this basis,an optimization model of individual bus line departure schedules based on dynamic demand is established,genetic algorithm is used to optimize the solution,and the influence of decision preferences on timetable is studied.The validity of the timetable optimization model is verified by a case study of the bus line 2 in the urban area of Weichang County.