Research On Travel Characteristics Of Beijing Rail Transit Based On Association Rules Mining

With the rapid advancement of urbanization,the urban radius continues to expand.The travel distance of residents has been enlarged and the phenomenon of separation of work and residence has become more common.The demand for fast and convenient travel is heating.Meanwhile,with rapidly growing number of motor vehicles,urban traffic congestion has become more and more serious.As a metropolis,Beijing has a particularly serious problem of separation of work and residence and traffic congestion.In this context,urban rail transit has developed rapidly over the past few decades due to its large capacity,fast,safe,punctual,and low average energy consumption.It has gradually connected the major sectors within the city and built the framework of the city.In the era of big data,with the development of computer and data technology,traffic data and other geographic information data related to residents’ travel have exploded.The method of using big data to study residents’ travel behavior has become one of the mainstream models in the field of behavior research.AFC card swipe data in rail transit system has the characteristics of low cost,large amount of information,objectivity,and strong temporal and spatial continuity,which can help us efficiently study the characteristics of rail transit travel.In order to study and analyze urban rail transit from multiple dimensions such as time,space,and routes,and find new ideas for understanding rail transit passenger flow,Association rule mining method is introduced in this paper.Association rule mining is a classic data mining method.It first finds out the frequent itemsets that often appear in the data set,and then finds the correlation between the frequent itemsets,discovers some interesting phenomena,and further searches for the existence behind this correlation.The potential causality of these phenomena can be reasonably explained.Different from traditional statistical analysis,Association rule mining can dig out the unknown mechanism behind common phenomena,and can also dig out the underlying laws of rare phenomena.Using the Beijing rail transit AFC card swiping data,the paper extracts the station information,time information and route information of rail transit trips,and uses the association rule mining method to study the single trip behavior.In terms of time,space,and route dimensions,single-dimensional analysis studies the characteristics of passenger flow and passenger travel behavior of Beijing rail transit,and then crossdimensional deeper analysis.The study found that commuting travel occupies an important position in Beijing’s rail transit travel,not only forming morning and evening peaks,but also spawning popular sites and popular routes;single large commercial sitemultiple small residential sites and a single large residential site-The grouping of multiple small commercial stations is more common,forming a group with obvious passenger flow direction;several small stations such as Tiananmen West Station have concentrated travel during off-peak hours,and there may be small groups with special travel needs;travel on the same route accounted for 35.8% of all trips,and suburban routes have a strong dependence on the main routes connected to it;63 large passenger flow stations bear 50% of the passenger flow of the entire network.The paper provides new ideas for understanding urban rail transit travel,and a new perspective for the study of urban rail transit passenger flow.It can help to carry out the analysis of the resource balance of rail transit system,assists in adjusting marketing strategies,maintains frequent passengers,and improves rail transit the performance of the transportation system.