Research Of Urban Subway Traffic Pattern Based On Big Data Platform

With the rapid development of economy and the constantly increasing of urban population, the city is expanding constantly, and as urban transportation the subway network plays a more and more important role. The movement of individual in the space is seemingly random and without regular, however, for the movement of a larger population such as the movement of crowd in the subway network, there is a specific model of the crowd. In order to study the intelligent transportation, spread of infectious diseases and other problems, based on individual behavior mobility, human movement mode can be mined. As a kind of complex network, transportation network can be used to study human mobility. So for certain patterns hidden in large population movement, the research of traffic network such as the subway network can be used to excavate urban subway traffic patterns, which can support the city planning, transportation planning, public health, social networks and other corresponding decisions. Therefore, the research of the urban subway traffic patterns is crucial.With the development of society and Internet and the advent of cloud age, big data gets people’s attention, and the research and application of the large data is increasingly mature. The basis of the research of transportation networks such as the subway network is the traffic data and traffic data collection method and processing method. With the development of the urban intelligent traffic system, new tracking technology such as global positioning systems, traffic sensors, etc. is used to trace human movement in big cities by high resolution, which also will inevitably produce a massive traffic data. For the big city traffic data, big data platform can be used for its processing and computing.This paper is mainly aimed at Shanghai traffic data. Based on the Hadoop big data platform, Shanghai citizen ID card data can be processed. This paper has set up a Hadoop big data platform, Hadoop standalone/installed pseudo-distributed configuration as well as the cluster configuration, and can look up the node information and tracking tasks in progress on the Web. Put the traffic data on HDFS system to analysis and process. The Hadoop big data platform has very strong computing ability, and can nicely process the big scale experimental data of Shanghai citizen ID card data.This paper first constructs the subway station passengers contact matrix, which express the passenger flow interaction between Shanghai 313 stations. According to the matrix, the community analysis and behavior stability analysis are proceeded, namely community detection and behavior entropy measurement. In the view of community analysis, different from intuitive geographic areas in Shanghai, according to the passenger flow, redefine the new concept of community, and it expresses the state of the community. Community detection very important for understanding complex network structure and obtain useful information or pattern. In this paper, the classic community discovery algorithm is used to divide Shanghai subway stations into different communities, and mine interactive pattern between communities. The internal link in community is tight, and external link is sparse. From the view of subway stations, several subway lines may be in the same community. From the view of region, several regions may be associated with the same community. It can be applied to functional area detection, early warning detection, etc. In the view of behavior analysis, this paper is mainly based on the subway passenger statistic to analyze its stability behavior. In this paper, the behavior entropy is used to measure human behavior. Based on Hadoop big data platform Shanghai passenger information entropy is calculated. It can be found that the value of passenger behavior entropy has certain periodicity, namely, the stability of passenger behavior has certain periodicity. These two aspects of the analysis results play a vital role in corresponding decision-making such as city planning, transportation planning, etc.