数据融合技术在高速公路异常事件检测中的应用

交通异常事件的检测一直以来都是高速公路监控系统的难点和重点。现有的交通异常事件检测系统在实际应用中检测效果不很理想。而数据融合技术是近年来发展起来的一种新的信息综合处理方法，目前已经在ITS中有许多尝试。本文简述了交通异常事件检测的基本原理，并通过对数据融合基本理论的总结与剖析，重点阐述了数据融合的3个层次（数据级融合、特征级融合、决策级融合）在交通异常检测中的具体应用以及交通异常事件检测的数据融合系统模型。应用数据融合技术的交通事件检测系统能提高系统的有效性，得到最佳协同作用的结果。

Application of Data Fusion Technique in Freeway Traffic Incident Detection

The detection of traffic incident has all being difficulty and significant in the monitoring and controlling system of freeway all the time. Existing traffic incident detection system is unsatisfactory in practical application. Data fusion technology is a new kind of information comprehensive processing method in recent years, there are already a lot of tries in ITS at present. This paper sketches the basic principle of traffic incident detection, and summarizing and analyzing the basic theories of data fusion, it has especially explained three levels of data fusion in application of traffic incident detection. This paper also illuminates the data fusion systematic model of traffic incident detection. The traffic incident detection system that uses the data fusion technology can improve systematic validity and get the optimal result in coordination with function.