Research On Station Identification Algorithm Based On Public Transport Big Data

With the rapid economic development in China,traffic jam in cities is growing.Advocating the public transportation is the best solution to traffic jam.With the accumulation of IC card data,researchers could find the behavior pattern of passengers and provide effective information for urban construction and the development of public transportation system.However,the IC card data only record the boarding time of passengers,which are seriously incomplete.Aiming at this problem,this paper analyzes the citizens' traveling behavior to studing station passenger flow and the identification of passengers' boarding/alighting stations.The outline of this research is as follows:Firstly,aiming at improving the boarding station identification algorithm based on passengers' transfer behavior,this paper analyzes the passengers' traveling behavior with IC card,then proposes the boarding station identification algorithm based on passengers' behavior.The new algorithm contains two innovations.One is that there are multiple cross points in the process of transfer within two lines.In order to solve this problem,this new algorithm uses metro-bus transfer information to identify the boarding stations of some passenger groups,then cuts IC card records and bus stations into several segments.In addition,in each segment,filtering the candidate stations of each bus transfer passenger according to the sequential relationship,we could get the boarding stations of passenger groups with bus transfer information.The other is that there are less transfer infornation about the passengers in morning rush hours and the time before it,which caused the invalid identification in passengers' boarding station.In order to solve this problem,this new algorithm finds the boarding stations in target lines of IC card loyal passengers during working days,which is helpful to identify the boarding stations of other passengers.Secondly,this paper proposes the alighting station identification algorithm based on the big data to solve the problem of the alighting station identification algorithm based on probability.With the basis of current theories,considering the station boarding amount in working days and characteristics of passenger flow in different period,this paper improves the calculation method of station attraction coefficient to get more objective and stable station attracting coefficient,the passenger flow and the characteristics of the passenger flow.The experimental data of this research are all from the Chongqing public transport big data.This paper compares accuracy and the average relative error between the present algorithm and the former algorithm.The results show that the algorithm proposed in this paper has better effects on the identification of boarding station and alighting station.