Cluster Analysis And Application Of Beijing Urban Rail Transit Stations Based On Passenger Flow Characteristics Data Of POI

As the domestic urban rail transit network becomes more and more complex,scientific and reasonable station type identification has an important auxiliary role for urban rail transit network planning,passenger flow forecasting,and station operation management.In order to guide the planning of the space around the station,to guide the passenger flow management organization of the station in a more targeted manner,and to provide a more scientific basis for passenger flow prediction,it is necessary to classify the station.In this paper,the original AFC card swiping data of urban rail transit in Beijing is cleaned,and the data is integrated into the conventional index of urban rail transit and station level evaluation index;the POI data of Beijing city is reclassified and sorted,and the quantitative The check index of the system;a buffer zone centered on the station was generated as the research object.This paper divides the date into three categories: working days,weekends,and holidays,and analyzes the characteristics of passenger flow based on the data.It analyzes the characteristics of one-week passenger flow,single-day passenger flow,and peak hourly passenger flow.It also analyzes the line passenger flow characteristics and cross section.Passenger flow characteristics.In this paper,based on the clustering of passenger flow characteristics time series,seven clustering variable indexes are constructed,including station attribute indexes,all-day timesharing passenger flow indexes,all-day passenger flow indexes,monthly passenger flow indexes,station traffic characteristic indexes,passenger flow Feature attribute index,project scoring index.Using standardized methods and principal component analysis to reduce the dimensionality of multi-dimensional variables,and compared the clustering effect before and after processing.By comparing the K-MEANS algorithm,hierarchical clustering algorithm,DBSCAN algorithm,mixed Gaussian model algorithm,it is decided to use the improved KMEANS algorithm to cluster the stations.The optimal k value was determined by the elbow method and the contour coefficient.This article also draws on the idea of quantitative identification of urban functional areas,uses Arc GIS tools to organize urban POI data into verification indicators,and constructs cluster verification indicators,including density verification indicators,distance verification indicators,and location verification.Index,frequency density index,category ratio index.This paper takes Beijing urban rail transit as an example,and divides 328 stations into 6 categories,namely,transportation hub station,residential area station,residential and employment dislocation mixed zone station,residential-based mixed zone station,and employment-based station Mixed zone site,residential employment and entertainment mixed zone site.After classifying the stations,suggestions for optimization of traffic organization,passenger flow management,and marketing organization are proposed for each type of station.This paper includes 44 pictures,39 tables,and 34 references.