Research On The Effect Of Time Differential Pricing Of Beijing Subway On Passenger Travel Behavior

With the rapid development of metro construction in major cities,the role of subway in the urban public transportation system has become increasingly important.Followed was the peak congestion problem caused by unevenly distributed passenger flow.Thus,government has adopted a series of measures to manage and adjust the demand,including time differential pricing which use the leverage of prices.Beijing Subway took the lead in adopting time-sharing pricing on some lines at the end of 2015 in order to influence passengers' travel behavior and re-distribute the passenger.Therefore,it is of great significance to study the impact of time differential pricing policies on passengers.Firstly,the implementation of time differential pricing policies at home and abroad was elaborated,and the current research situation was analyzed.The basic theories involved in the study of time differential pricing,including crowd pricing,cluster analysis,SP survey methods and the Logit model were introduced.Secondly,the specific measures for time differential pricing in Beijing Metro was analyzed.In order to analyze the impact of the policy in depth,clustering based on AFC data was used.The clustering process classifies Beijing subway passengers into four categories.The characteristics of each type of passenger were analyzed.Based on the classification results,the effect of pre-peak concession fares on each type of passengers was measured.Thirdly,in order to explore the impact of pre-peak concession fare on passengers'behavior,4374 pieces of data were obtained from the SP survey.Through the cross-analysis of individual attributes,the conclusion was drawn that there were differences among the passenger groups,and it was proved that while analyzing the effect of pre-peak concession fares on passenger travel behavior,it is necessary to analyze based on passenger classification.Fourthly,multinomial logit model was used to curve passenger choice behavior in the condition of time differential prcing based on passenger classification.Comparisons were made with models without classification.The Logit model based on passenger classification can recognize the differences in behavioral influence factors,including the similarities and differences between each type of passengers' sensitivity to discounts,crowding levels.Furthermore,the socioeconomic attributes of passengers were included into the model through cross-items,taking the impact of individual attributes on passengers' selection behavior into account.It was found that income and travel time is the major factors affecting travel time selection under discounted fare conditions;finally,the model is used to measure the relationship between the proportion of transferred passengers under different fare discount levels and transfer time.The results show that the pre-peak concession fare has a certain significant effect on the passenger flow within thirty minutes after the critical time,and has a general effect on passenger flow transfer beyond thirty minutes.