交通运输系统工程与信息

交通运输系统工程与信息 ›› 2019, Vol. 19 ›› Issue (2): 66-72.

• 智能交通系统与信息技术 • 上一篇    下一篇

基于卡口车牌识别数据的车辆出行分析

龙小强 1, 2，苏跃江*1, 2，余畅 1, 2，吴德馨 1, 2   

  1. 1. 广州市交通运输研究所，广州 510000；2. 广州市公共交通研究中心，广州 510000
  • 收稿日期:2018-10-17 修回日期:2018-12-10 出版日期:2019-04-25 发布日期:2019-04-25
  • 作者简介:龙小强(1976-)，男，江西吉安人，高级工程师，博士.

Analyzing Methods of Vehicle' Travel Using Plate Recognition Data

LONG Xiao-qiang1, 2, SU Yue-jiang1, 2, YU Chang1, 2, WU De-xing1, 2   

  1. 1. Guangzhou Transport Research Institute, Guangzhou 510000, China; 2. Guangzhou Public Transport Research Center, Guangzhou 510000, China
  • Received:2018-10-17 Revised:2018-12-10 Online:2019-04-25 Published:2019-04-25

PDF

摘要：

车辆出行是城市道路交通的基本组成单元，掌握城市道路网车辆的出行信息，深入挖掘车辆出行特征与规律，能为城市交通管理提供决策信息.本文基于卡口车牌识别数据，提出了一套车辆出行分析框架.首先对全路网运行的所有车辆的个体出行进行辨识，提取所有车辆出行的路径和行程信息，并从个体和集计层面获取车辆出行的规律特征；利用车辆的多日出行信息和统计特征，提出了车辆职住地识别方法；基于外地车的出行特征，利用 K-means++ 算法对外地车进行分类.在实例分析中，以广州市道路网运行车辆作为研究对象，开展了车辆出行分析，实验结果验证了本文方法的有效性.通过本文方法挖掘的信息对城市道路交通管理具有重要意义.

关键词: 智能交通, 车牌识别数据, 个体出行信息, 出行特征, 职住地识别, 特征聚类

Abstract:

Vehicle travel is the basic component of urban road traffic. There is an urgent need for urban traffic manager to obtain the trip information on urban road network, and make a better understanding of vehicles' travel characteristics. In this paper, an analysis framework based on plate recognition data is proposed. Firstly, the individual trip information of all vehicles on urban road network is identified and the trip information of all vehicles are extracted. And then vehicles' travel characteristics are obtained at the individual and aggregate level. A methodology is proposed for recognizing vehicles' residence and workplace based on trip information and statistical feature. K-means ++ algorithm is applied to cluster and classify the nonlocal vehicles. In case study, focused on the vehicles of Guangzhou City, the analysis of vehicles travel are conducted. The effectiveness of the proposed method is proved by the case study results. The information mined through the method is of great significant for urban road traffic management.

Key words: intelligent transportation, plate recognition data, individual trip information, travel characteristics, residence and workplace identified, feature clustering.

中图分类号: