Research On The Correlation Between Passenger Travel Characteristics And Ticket Fare Pricing Of Urban Rail Transit

With the rapid economic and transportational development,society enters a new stage of urbanization and motorization.what's more,traffic congestions has become increasingly problematic because of huge traffic demand.Compared to other mode of transport,rapid transit stands out with high capacity,safety,eco-friendly efficiency.Therefore,investigation of travel connection between passengers and fares through big data of subway passenger flow is of great significance to deal with current traffic congestion issue.This will provide guidance for fare adjustment as well as the balance of citizen's travel,which would ultimately improve subway service,alleviate urban congestion,and further promote urban rail transit development.First,with the help of big data analysis,this paper extracts and analyzes macro travel characteristics of subway passengers,namely the frequency,time,and location of card swiping,to capture peak hour and major station information for rapid transit traffic.Then,passenger behavior is further studied with K-means algorithm in three dimensions: travel time,travel frequency and travel expense.Specifically,cluster passengers into several types based on a few features(number of trips,proportion of peak trips,proportion of off-peak trips,travel expenses,proportion of expense in peak hours and off-peak hours,etc)and summarize passenger behavior preference for each type so as to quantify their travel intention.Finally,a fair fare charging model that satisfies passenger behaviors can be established by the big data analysis method,with the best peek fare increasing rate,non-peak fare decreasing rate and adjustment factor introduced to improve passenger transfer intention.Take subway data as an example,this paper implements K-means clustering algorithm to analyze passengers' travel behaviors through tools such as python,Matlab,SPSS,etc,which would enhance the urban rail transit system with more differentiated fares.With the help of differentiated fare model generated from one month card swiping data,this article proposes the best price adjustment plan that would effectively benefit enterprises,government and passengers: raise fare price by 35% during peak hour while reduce by 54.7% during off-peak hour.