Spatial-temporal Analysis Of Urban Public Transportation Big Data Based On Hadoop

Urban public transportation big data contains rich spatio-temporal information,which is the data foundation for passenger travel characteristics analysis,urban transportation service capacity evaluation and public passenger flow prediction.It is the data foundation for passenger travel characteristics analysis,urban transportation service capacity evaluation and public passenger flow prediction.It is also an important basis for urban public transportation scientific management and planning and design.With the rapid development of urban transportation,urban management departments have accumulated large-scale transportation data.However,these data cannot be fully utilized,because traditional traffic data analysis software cannot load,process or analyze the big data,when the method of extracting data from big data is generally used,the sample size is small,and the analysis results are mostly based on two-dimensional charts,so the data visualization effect is generally.In order to solve this problem,this article takes bus card swipe data and taxi GPS trajectory data in beijing as the research object,the main research content of this article are:(1)Build a Hadoop distributed computing platform.In this article,eight computers in the laboratory are used to build a micro-distributed cluster,30 days of bus card swipe data and7 days of taxi GPS trajectory data are distributed and stored,and the MapReduce programming model is used for data cleaning,preprocessing and mining.(2)Build bus passenger travel chain model and vehicle passenger flow model.This article establishes a bus passenger travel chain based on space-time constraints,a bus passenger flow model based on bus stops,and a taxi passenger flow model based on vehicle status.Based on the above model,the bus interchange data,bus passenger flow data,and taxi passenger flow data are extracted and calculted.(3)Spatial-temporal analysis of urban public transportation big data.According to the operating characteristics of different types of vehicles and passenger travel characteristics,a variety of urban public transportation big data analysis methods including bus line operation analysis,transportation passenger flow analysis and identification of passengers' residence are designed and implemented,and the difference analysis of the travel characteristics of passengers of different types of transportation is carried out from three aspects: travel time,travel hotspot and travel time.(4)Data visualization based on WebGIS.In order to fully demonstrate spatial-temporal characteristics of traffic big data,the article integrates online map services,2D spatial data and 3D model data,and uses multiple expressions to visualize the calculation results of traffic big data analysis,such as heat maps,3D histograms,interactive maps and non-spatial geographic charts.(5)Integrate WebGIS and Hadoop distributed computing platform in technology.In order to make full use of the rich expressions of WebGIS and the advantages of Hadoop distributed storage and computing,this article implements closed-loop process operations from data request issued by the client,distributed computing of Hadoop cluster to datavisualization using WebGIS.This article builds a Hadoop distributed computing platform for traffic big data mining,which lays a solid foundation for comprehensive and multi-angle analysis and calculation of urban public transportation big data,then develops a WebGIS-based visualization system to visualize the results of data analysis,which brings people a more intuitive visual experience and is helpful for discovering the hidden rules of visualization,then integrates WebGIS and Hadoop,which does not require manual midway intervention,so as to solve the problem of centralized call and display of big data.