Metro OD Matching and Statistical Research Based on AFC Data

作者：Ailin Jiang, Qianqian Sun, Zhen Peng, Xinyi Wei, Qiang Huang, Tong Ming · 发表于：Journal of Physics Conference Series · 年份：2020 · DOI：10.1088/1742-6596/1624/3/032051 · 被引用次数：2 · 研究领域：Traffic Prediction and Management Techniques、Transportation Planning and Optimization、Human Mobility and Location-Based Analysis

Abstract With the development of the smart city system, it is more and more import to improve China’s rail transit system and create a convenient traffic environment for passengers. This article will use the AFC data of Chongqing Regional Rail Transit for two months for OD(origin-destination) matching and passenger flow statistics. The main work includes AFC(Automatic Fare Collection) data cleaning, OD information matching, data conversion, passenger flow statistics and data visualization under the big data environment. At the same time, in the construction of the data platform, three schemes are currently adopted. One is to adopt the traditional method of java-Oracle when the amount of data is small. The second is to use the classic method of Hadoop-Hive when the amount is large. The third is to use parallel computing Maxcompute for large amounts of data to improve data calculation speed. Through the experimental comparison of the three schemes, the exact matching of the site, line, time and passenger flow statistics, in terms of speed, the parallel computing method is used to calculate the fastest data. This paper based on the AFC data of metro OD matching and statistical research can help the transformation of urban traffic structure from extensive to fine and provide support for urban transportation planning and decision-making.