基于多源信息融合的高速公路事件检测算法研究  

Freeway Incident Detection Algorithm Based on Multi-source Information Fusion

作　　者：陈扶崑[1] 吴中[1] 田亮[1] 

机构地区：[1]河海大学,南京210098

出　　处：《交通信息与安全》2009年第1期35-38,共4页Journal of Transport Information and Safety

摘　　要：为了研究高速公路交通事件检测算法,以多源信息融合理论为基础,依托人工神经网络技术,设计了固定检测器与浮动车检测器的信息融合事件检测算法,并说明了具体的检测原理和融合过程。通过Vissim仿真获得数据,在Matlab中编程实现了信息融合过程,试验结果表明在三级报警策略下,信息融合算法的事件检测率、误报警率和平均检测时间都达到了较高的检测水平,证明了所设计的信息融合交通事件检测方法的优越性。To study the detection algorithm of freeway traffic incident, this paper designed an information fusion detection algorithm of fixed and probe vehicles detectors based on multi-source information fusion theory and artificial neural network technology. It described the specific principles and integration process. Finally, this paper obtained the data by VISSIM and realized the process of information fusion in Matlab. The test results show that the incident detection rate, false alarm rate and mean detection time have reached a relatively high level in three-stage alarm strategy. Therefore, it proves the superiority of the information fusion traffic incident detection method.

关 键 词：自动交通事件检测 信息融合 人工神经网络 浮动车 VISSIM 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]