摘要
交通流状态分类对于选择交通控制和诱导策略有非常重要的作用,不同的快速路路段设定的交通流参数临界值及变化特性会有所不同.本文考虑到交通流参数对交通流状态判别的影响程度,给出了一种基于加权欧氏距离的相似性度量方法,并确定了交通流状态判别的关键参数.根据整个路段的交通流数据,通过聚类分析构造最小距离分类器,把个别路段的交通流数据作为样本数据,进行了对个别路段的状态评估.实证分析结果表明:在交通流状态判别过程中,密度是最关键的参数:基于最小距离分类的个别路段的状态评估结果与实际情况非常类似,这将为交通控制和管理提供决策依据.
Classification of the traffic flow state plays a very important role in the traffic control and guidance, there are difference of threshold of the traffic flow parameters in different freeway road section. In this paper, it is considered that each parameter has influence on judgment of the traffic flow state, a similarity measurement based on weighted Euclidean distance is proposed, and the most important parameter for state judgment is determined. Minimum-distance classifier is built by clustering analysis using the traffic flow data for whole road section, and then the traffic flow states of individual road section are estimated. The experimental analysis shows that density is a very important parameter for judgment of the traffic flow state; the experimental results using minimum-distance classification are similar with the real situation. The results can be useful for traffic control and management.
出处
《北京交通大学学报》
CAS
CSCD
北大核心
2009年第6期47-51,共5页
JOURNAL OF BEIJING JIAOTONG UNIVERSITY
基金
国家自然科学基金资助项目(60874078)
新世纪优秀人才支持计划项目资助(NCET-08-0718)
高等学校博士学科点专项科研基金项目资助(20070004020)
关键词
交通流状态
交通流参数
加权欧氏距离
最小距离分类
traffic flow state
traffic flow parameter
weighted-Euclidean distance
minimum-distance classification