考虑多源不确定信息的路网交通拥堵状态辨识方法

拥堵状态辨识是道路运行状态评估的重要内容，是交通系统流量调控和管理的重要参考指标.在智能交通系统（Intelligent transport system，ITS）普及化程度越来越高的后交通时代，如何实现海量数据下对多源不确定交通拥堵状态的辨识是非常重要的内容.首先，基于多元集对分析建立一种新的路网交通拥堵状态刻画模型；然后，通过改进证据理论中Dempster组合规则实现交通信息融合，并推导出当前交通拥堵状态的准确表达值；最后，在数值模拟的基础上，使用重庆市南岸区的交通检测数据进行仿真分析，结果表明本方法能准确直观地反映出实时交通拥堵状态，具有潜在的实际应用价值.

Traffic Congestion Status Identification Method for Road Network with Multi-source Uncertain Information

Congestion identification is an important content of traffic condition assessment, and has significant meaning to the traffic regulation and management of transportation systems. With intelligent transport system (ITS) becoming increasingly popular, how to achieve congestion identification for uncertain multi-source information is a very important content under massive data. First, a new road network traffic congestion state characterization model is built based on multivariate set pair analysis method. Then, traffic information fusion is achieved by improving the Dempster combination rule of evidence method, and the accurate expression values of current traffic congestion are derived. Finally, the real time traffic monitoring data in Chongqing is used to verify the presented method. The results illustrate the presented method is effective, and that it is not only of theoretical significance but also of potential application value.