基于决策树C5.0算法的城市非法营运客车识别预测  

Identification and Prediction of Unlicensed Passenger Cars in Urban Areas Based on the Decision Tree C5.0 Algorithm

作　　者：肖赟 李金艳 李荣巧 Xiao Yun;Li Jinyan;Li Rongqiao(School of Urban Construction and Transportation,Hefei University,Hefei,Anhui 230601,China;Anhui Provincial Intelligent Transportation Big Data Analysis and Application Engineering Laboratory,Hefei,Anhui 230601,China;School of Civil Engineering and Transportation,Northeast Forestry University,Harbin,Heilongjiang 150040,China)

机构地区：[1]合肥大学城市建设与交通学院,安徽合肥230601 [2]安徽省智慧交通大数据分析与应用工程实验室,安徽合肥230601 [3]东北林业大学土木与交通学院,黑龙江哈尔滨150040

出　　处：《黑龙江工业学院学报(综合版)》2024年第12期99-103,共5页Journal of Heilongjiang University of Technology(Comprehensive Edition)

基　　金：安徽省高校自然科学基金重点项目“基于车辆路径特征的非法营运行为智能甄别和数字化治理方法研究”(项目编号:2022AH051800)。

摘　　要：针对非法营运客车识别和治理费时费力、精度不高问题,基于交通卡口数据,设计非法营运客车识别多维特征指标,并利用决策树C5.0算法构建识别模型。以安徽省某市为研究对象,采集该市城区交通卡口车辆通行记录数据,结合出租车、已查实非法营运客车信息,构建识别模型以预测疑似非法营运客车。结果表明,识别模型精度为94.87%,能够有效识别疑似非法营运客车信息,可为实际非法营运客车识别治理提供有效方法支撑。Regarding the problems of time-consuming,laborious and low accuracy in the identification and governance of unlicensed operated cars,based on traffic bayonet data,multi-dimensional characteristic indicators for the identification of unlicensed operated cars are designed,and an identification model is constructed by utilizing the Decision Tree C5.0 algorithm.Taking a city in Anhui Province as the research object,the vehicle passing record data of traffic bayonets in the urban area of this city are collected.Combined with the information of taxis and verified unlicensed operated cars,an identification model is constructed to predict the suspected unlicensed taxis.The results show that the accuracy of the identification model is 94.87%,which can effectively identify the information of suspected unlicensed taxis and can provide effective methodological support for the actual identification and governance of unlicensed taxis.

关 键 词：交通工程 非法营运客车 决策树算法 预测模型 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] U491.2[自动化与计算机技术—计算机科学与技术]