Analysis And Modeling Of Microscopic Fluctuation Of Passenger Flow In Scheduled Transportation System

Scheduled transportation system,oppositing to the concept of free transportation system,has become an indispensable part of people’s daily travel.However,there is still a lack of specific and systematic research on the characteristics of passenger flow in scheduled transportation systems.The microscopic fluctuation characteristic is also an important feature of scheduled transportation systems.In view of this,the research content of passenger flow fluctuation characteristics of the scheduled transportation system is as follows:First of all,based on the scheduled transportation systems in life —— the urban rail transit system,the ground bus system,the railway passenger transport system and the civil aviation system,the operational characteristics and passenger flow characteristics of the scheduled transportation system are summarized.Operational characteristics include:(1)with timetable;(2)independence;(3)there are different numbers of fixed lines and fixed sites in the system;(4)transport capabilities include both pass ability and delivery capability.The characteristics of time distribution of passenger flow include:(1)discontinuous flow characteristic;(2)significant periodicity;(3)microscopic fluctuation characteristics.Spatial distribution characteristics include:(1)uneven distribution between the lines;(2)uneven distribution between the sites;(3)uneven distribution in a site.Secondly,using passenger flow data of Suzhou Rail Transit and Xuzhou Bus No.1 as the example data.After preprocessing the original data,the passenger flow fluctuations are analyzed preliminarily.It can be found that,after the passengers are grouped according to the time period,the trend of the passenger flow within the group is periodic;the mean value and variance obtained by grouping show different distribution trends within a day.Then,three methods to model and calculate the fluctuation characteristics of passenger flow in scheduled transportation system are proposed:(1)continuous passenger flow method;(2)classification method;(3)period grouping method.These three methods are used to model the passenger flow fluctuations of the Suzhou rail transit passenger flow and the Xuzhou Bus No.1 passenger flow at the 5min,15 min,and 1h aggregation levels respectively.And,passenger flow fluctuation value and passenger flow fluctuation intervals are calculated.Based on this,according to the principles of modeling and calculation,a GUI for predicting passenger flow fluctuation in scheduled transportation system is formed.Finally,mean absolute error(MAE),mean absolute percentage error(MAPE)and kickoff percentage(KP)are used to evaluate the performance of the modeling and calculating methods.Width-to-flow ratio(R)is used to compare and analysis fluctuation characteristics of passenger flow.Comparing three modeling methods,it is found that,the calculation results of workday passenger flow fluctuation value obtained by the classification method are the most accurate,and the calculation result of passenger flow fluctuation interval obtained by the period grouping method is most consistent with the 95% confidence level.A comparative analysis of results under different aggregation levels shows that,when the time aggregation is larger,the calculation accuracy of the passenger flow fluctuation value is higher.And,when the time aggregation is smaller,the calculation result of the passenger flow fluctuation interval is more consistent with the 95% confidence level.The fluctuation characteristic of passenger flow within a day is that,during the morning and evening peak hours,the fluctuation is weaker.The impact of time aggregation on the fluctuation of passenger flow is that,the fluctuation of passenger flow decreases with the increase of time aggregation.The impact of workday and rest day on the fluctuation of passenger flow is that,the fluctuation of passenger flow on workday during the peak period is stronger than that on rest day,and the fluctuation of passenger flow on rest day of other period is stronger than that on workday.