基于纵向时间序列的快速路交通事件检测算法

牛世峰¹, 姜桂艳^1,2, 李红伟¹, 姜卉¹

1.吉林大学交通学院;2.吉林大学汽车动态模拟国家重点实验室



摘要:

为了提高快速路交通事件检测算法的效率,减少交通事件对快速路交通流的影响,在提出交通参数数据纵向时间序列定义的基础上,重新界定了异常交通状态的概念,分析了异常交通状态与交通事件之间的关系,利用增益放大原理设计了交通状态变异指数和以此为基础的交通事件检测算法,在确定交通事件位置的同时,给出事件的空间影响范围,并以某城市快速路的实际交通数据为基础进行了对比分析.结果表明,所提出的交通事件检测算法的检测效果明显优于对比的经典算法,且具有更好的鲁棒性.

关键词:  快速路  交通事件  异常交通状态  自动检测

DOI：10.11918/j.issn.0367-6234.2011.02.029

分类号:U491.31

基金项目:国家高技术研究发展计划(2007AA11Z245);高等学校博士学科点专项基金(20070183129)



Automated detection algorithm for traffic incident in urban expressway based on lengthways time series of traffic parameters

NIU Shi-feng¹, JIANG Gui-yan^1,2, LI Hong-wei¹, JIANG Hui¹

1.Transportation and Traffic College,Jilin University,130022 Changchun,China;2.State Key Laboratory of Automotive Dynamic Simulation,Jilin University,130022 Changchun,China,)

Abstract:

In order to improve the efficiency of traffic incident detection algorithm in urban expressway and reduce the impact of incidents,basing on proposing the definition of lengthways time series of traffic parameters and defining abnormal traffic statuses clearly,the relation between abnormal traffic statuses and traffic incident is analyzed,the traffic status variability index is constructed using gain amplification principle,then automated detection algorithms for the traffic incident is designed.This method can not only detect the location of traffic incident but also the incidence of traffic incident,the result show that detection effect and robustness of this algorithms is better than classical algorithms.

Key words:  urban expressway  traffic incident  abnormal traffic status  automatic detection