Data-driven Assessment Of Passenger Flow Aggregation Degree Of Subway Network During Peak Hours

With the development of the city,the subway lines are continuously built and expanded rapidly to form a network structure.Because the subway has the characteristics of fast speed and punctuality,it undertakes most of the passenger flow transportation tasks in the city.The sharp increase in the number of commuter passengers during peak hours has led to problems such as high train load rates and a large number of passenger retention.The phenomenon of congestion on the station platform often occurs,with the network operation,and this state spreads between stations,which greatly increases the difficulty of subway safety management.Based on this,this paper proposes the dynamic aggregation degree of passenger flow to measure the congestion degree of subway station passenger flow and the distribution of network passenger flow aggregation.Macroscopically grasp the changing process of subway passenger flow aggregation state,and provide decision support for passenger flow management at the line network level.The main research work in this paper is as follows:(1)The basic theory of passenger flow aggregation in subway network system is studied from three aspects: the composition of subway network,the reasons and influencing factors of passenger flow aggregation,and the classification of aggregation degree.The analysis shows that the lack of transportation capacity of the line is the main reason for the accumulation and retention of passengers on the platform,and the network effect is the way of the aggregation and spread of the passenger flow.Therefore,the passenger flow aggregation degree of the subway network is described by combining these two aspects.(2)The calculation model of the number of people in the subway network is constructed.Based on the card swiping data,the travel time composition of passengers is analyzed,and the Dijkstra algorithm is used to determine the route selection of passengers.Match the train running timetable,extract the passenger data of only one train that can take the train,and estimate the platform travel time.This paper clarifies the calculation method of the passenger retention time in the itinerary data,and proposes a method to calculate the aggregated number of network passengers in different time granularities.(3)Three centrality indexes of the integrated network topology structure measure the structural importance of stations in the subway network,and propose a calculation method for the dynamic aggregation degree of the subway network combined with the station dynamic passenger flow aggregation data.The aggregation status of online passenger flow is divided into four levels: "no aggregation,slight aggregation,moderate aggregation,and severe aggregation".(4)Take the passenger flow data of Shanghai subway network as an example to analyze.The temporal and spatial changes of the total number of passengers and the number of passengers retained on the Shanghai subway network and stations are analyzed in a refined time granularity of 1 minute.Calculate the dynamic aggregation degree of network passenger flow,and use K-means clustering algorithm to calculate the cluster center of passenger flow aggregation degree at station and network level respectively.Among them,the station cluster centers are(0.046,0.177,0.415,0.783),and the clustering degree classification range thresholds are(0-1.12,1.12-2.96,2.96-5.99,5.99-10).The cluster center value of the passenger flow aggregation degree of the subway network is(1.96,4.04,6.62,5.58),and the threshold value of the aggregation degree classification range is(0-3,3-5.33,5.33-6.1,6.1-10).The final analysis verifies the rationality of the aggregation degree of network passenger flow.