基于数据融合的高速公路交通状况检测

发布时间：2018-01-11 15:06

本文关键词：基于数据融合的高速公路交通状况检测　出处：《计算机学报》2017年08期 　论文类型：期刊论文

【摘要】：物联网的普及使得数据感知和共享变得迅速、便捷,充分挖掘和利用这些数据成为当前物联网研究的热点.就高速公路交通状况检测而言,有效使用收集到的数据需要应对两个挑战:一是单一感知数据内在的稀疏性;二是检测需求的多目标性.为了高精度、低成本、全路段、准实时地检测高速公路交通状况,该文依托移动通信网的信令数据,结合车辆检测器数据,提出了一种数据融合方法即Megrez方法.Megrez方法不仅挖掘了感知数据的内在特性以重构缺失数据,而且在对重构数据进行纠偏时融入了交通流的特征.由于该方法以非介入方式判定交通状况,因此,能以极低成本实现高速公路全路段的路况检测,同时,在并行线性代数库支持下实现的Megrez方法能在指定时间片内完成对车速的估计.该文基于大量真实数据从不同角度对Megrez方法进行了评估,实验结果表明Megrez方法估计的车速具有较高的准确性,能准确检测高速公路的交通状况.
[Abstract]:The popularity of the Internet of things makes data sharing become aware and rapid, convenient, make full use of these data has become a hot topic in the research on the Internet of things. The detection of expressway traffic condition, the effective use of the collected data need to deal with two challenges: one is the single sensing data sparsity inherent; two is multi target detection in order to demand. High precision, low cost, full sections, quasi real time detection of expressway traffic situation, this paper based on the signaling data of mobile communication networks, combined with the vehicle detector data, put forward a method of data fusion method named Megrez.Megrez mining method not only the intrinsic characteristics of the sensing data to reconstruct the missing data, and into the the characteristics of traffic flow in correction of the reconstructed data. Because the method to determine non intervention traffic conditions, therefore, to the very low cost to realize the highway The road traffic detection, at the same time, the parallel Megrez method of linear algebra library support to achieve the estimation of speed can be completed within the specified time in this paper. Based on a large number of real data from different angles on the Megrez method were evaluated, the experimental results show that the Megrez method to estimate the speed with high accuracy, can accurately detect the highway the traffic.

【作者单位】：中国科学院软件研究所计算机科学国家重点实验室;中国科学院大学;
【基金】：国家自然科学基金(61472408) 交通运输部项目(2015315Q16080)资助~~
【分类号】：TN929.5;TP202;U495
【正文快照】： 1引言随着物联网的普及,信息的获取和共享变得迅速和便捷[1],充分利用这些信息,将推动物联网应用向深度和广度发展.例如,在交通领域,借助于物联网提供的感知设备和网络,可以感知并收集到道路上与人、车、路、环境等相关的多种动静态数据.充分挖掘和利用这些数据,可以为大众提

【相似文献】