基于ETC收费流水数据的非法客运车辆特征研究  

Research on the Characteristics of Illegal Passenger Vehicles Based on ETC Toll Pipeline Data

作　　者：田钦瑞 李桥兴[1] 俞杰 杨城 苏锦玉 TIAN Qin-rui;LI Qiao-xing;YU Jie;YANG Cheng;SU Jin-yu(School of Management,Guizhou University,Guiyang 550000;Guizhou Zhitong Tianxia Information Technology Co.,Ltd.,Guiyang 550000,China)

机构地区：[1]贵州大学管理学院,贵州贵阳550000 [2]贵州智通天下信息技术有限公司,贵州贵阳550000

出　　处：《物流工程与管理》2023年第4期124-128,161,共6页Logistics Engineering and Management

基　　金：贵州省哲学社会科学规划联合基金课题“基于自组织理论的贵州省大数据产业生态系统演化机理研究”(18GZLH04)。

摘　　要：非法客运车辆的频繁通行影响了道路运输的安全和市场秩序的稳定。文中基于改进的K-means^(++)算法构建了非法客运车辆识别方法,并以贵州省贵阳市高速公路400万余条数据做实证研究,构建与通行天数等与时间相关的6个聚类特征变量,对约26万样本车辆进行聚类分析得到疑似非法客运车辆的分布情况,从车牌、时间和空间等方面获得疑似非法客运车辆的出行特征:车牌分布以贵阳本地车牌为主并占比约80%;非工作日客运车辆无明显进出站高峰期,而工作日客运车辆的进站高峰期为8-10点和出站高峰期为17-18点;车辆最常用的出行路线为贵州乌当站与贵州综保站间的往返路线;最大嫌疑车辆为在工作日9∶00-10∶00间进入贵阳市乌当收费站点的贵阳本地牌照车辆。研究结果既为贵阳市打击非法营运车辆提供可靠的分析依据,也为其他城市交通运输管理提供方法借鉴。The frequent traffic of illegal passenger vehicles affects the safety of road transportation and the stability of market order.In this paper,the identification method of illegal passenger vehicles is constructed based on the improved K-means^(++)algorithm,and based on more than 4 million data of expressways in Guiyang City,Guizhou Province,six clustering characteristic variables related to time,such as the number of days,are constructed,and the distribution of suspected illegal passenger vehicles is obtained by cluster analysis of about 260000 sample vehicles.The travel characteristics of suspected illegal passenger vehicles are obtained from the aspects of license plate,time and space:the license plate is mainly distributed in Guiyang and accounts for about 80%.There is no obvious peak period for passenger vehicles entering and leaving the station on non-working days,while the peak periods of entering and leaving stations for passenger vehicles on weekdays are 8-10∶00 and 17-18∶00 respectively.The most commonly used travel route for vehicles is the round-trip route between Guizhou Wudang Station and Guizhou Comprehensive Insurance Station.The most suspected vehicle is the Guiyang local license plate vehicle that enters the Wudang toll station of Guiyang City between 9∶00 and 10∶00 on working days.The research results not only provide a reliable analysis basis for the fight against illegal operating vehicles in Guiyang City,but also provide reference for other urban transportation management methods.

关 键 词：非法客运车辆 ETC收费流水数据 出行特征 聚类分析 K-means^(++) 

分 类 号：D922.296[政治法律—经济法学]