手机信令数据支持下高速公路检测器布设研究  

Freeway Detector Placement Strategy Research Based on Mobile Phone Handoff Data

作　　者：王浩淼 冉斌[2,3] 雷怡 杨彬彬 

机构地区：[1]上海美慧软件有限公司,上海200081 [2]东南大学物联网交通应用研究中心,南京210096 [3]威斯康星大学麦迪逊分校土木工程系,威斯康星麦迪逊53706

出　　处：《道路交通与安全》2016年第5期52-58,共7页Road Traffic & Safety

基　　金：交通运输部信息化技术研究项目<江苏省高速公路网运营与服务智能化平台关键技术深化研究>(项目编号:2015364X16030)

摘　　要：交通检测器的布设决定着路段行程时间估算及预测的精度.同时,检测器技术、行程时间估计方法也会影响到检测器布设方法.针对此问题研究利用手机切换数据交通信息提取支持,结合传统固定检测器,基于BP神经网络实现行程时间融合估计.基于新的行程时间估计方法,以估计精度为约束,研究传统固定检测器的布设.结合实例分析,验证行程时间融合估计方法有效性,并在均匀布设和盲区补充方案基础上,研究检测器布设间距对交通参数估计的影响.Freeway traffic detector placement is the prerequisite of freeway management operation and one of the important research subjects,which decides the accuracy of travel time estimation and prediction.Meanwhile,detection technologies and travel time estimation methodologies have a great influence on traffic detector placement. The research used the cellular handoff data to extract traffic information and combined with the data from traditional point detectors to realize travel time fusion estimation on the basis of the BP neural network. Based on the new estimation method,with estimation precision as constraints,the research focused on the traditional point detector placement. The effectiveness of travel time fusion estimation proposed got verified with a study case. Based on the uniform spacing strategy and blind area compaction strategy,the research explored the influence of the detector spacing on traffic parameters using the proposed travel time fusion estimation method.

关 键 词：智能交通系统 检测器布设 移动信令数据 BP神经网络 数据融合 

分 类 号：U491[交通运输工程—交通运输规划与管理]