基于信令流的高速公路车辆实时分类  

Exploiting signaling data stream to classify vehicles on highways in real time

作　　者：姬强 金蓓弘[1,2] 崔艳玲 张扶桑[1,2] JI Qiang;JIN Bei-hong;CUI Yan-ling;ZHANG Fu-sang(State Key Laboratory of Computer Sciences, Institute of Software, Chinese Academy of Sciences, Beijing 100190,China;University of Chinese Academy of Sciences, Beijing 100190, China)

机构地区：[1]中国科学院软件研究所计算机科学国家重点实验室,北京100190 [2]中国科学院大学,北京100190

出　　处：《计算机工程与设计》2018年第5期1439-1445,共7页Computer Engineering and Design

基　　金：国家自然科学基金项目(61472408)

摘　　要：为实现实时、全覆盖的高速公路车辆分类,提出基于手机信令数据的Lepus方法,其包括离线和在线两部分。在离线部分,根据同一时空下信令轨迹和带有车辆类别的GPS轨迹建立信令轨迹、车辆类别和信令可识别车辆之间的联系;基于这些标注后的信令可识别车辆,提取车辆分类的特征并用于分类模型的离线训练。在线部分,用分类模型处理到达的信令流,实现对车辆的实时分类。实验结果表明,Lepus方法可以有效实现实时的车辆分类。To achieve the real-time,full coverage classification of vehicles on highways,an approach named Lepus based on the signaling stream was proposed.The Lepus approach consisted of an offline part and an online part.For the offline part,by analyzing the historical GPS trajectories with vehicle types and the signaling trajectories occurring at the same time and space,the relations among signaling trajectories,vehicle types and signaling-recognizable vehicles were established.The driving characteristics of these labeled signaling-recognizable vehicles were analyzed to determine vehicle classification features and train a classification model.As for the online part,the incoming signaling stream was handled and the vehicles were classified in real time.Extensive experimental results show that the Lepus approach is effective in real time vehicle classification.

关 键 词：车辆分类 信令流 GPS轨迹 智能交通系统 移动感知 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]