详情

原文传递 EVALUATION OF INCIDENT DETECTION METHODOLOGIES.

题名：	EVALUATION OF INCIDENT DETECTION METHODOLOGIES.
作者：	Mahmassani-HS; Haas-C; Zhou-S; Peterman-J
关键词：	Algorithms-; Data-fusion; Field-studies; Freeways-; Incident-detection; Integrated-models; Interviewing-; Literature-reviews; Optimization-; Performance-; Technology-; Traffic-control; Traffic-data
摘要：	The detection of freeway incidents is an essential element of an area's traffic management system. Incidents need to be detected and handled as promptly as possible to minimize delay to the public. Various algorithms and detection technologies are examined to determine which combinations offer optimal detection performance. The objectives of this research are to compile, compare, rank, and recommend incident detection strategies in use today. Incident management and its components are described in this report to provide background. Extensive literature reviews, site visits, and interviews have been conducted and continue to be pursued as new incident detection schemes emerge. The most prevalent and practical incident detection algorithms are coded into software for testing and performance comparison. Large amounts of traffic data have been acquired for input into detection algorithms. An integrated incident detection data and algorithm fusion model is proposed as well. This model can be used both as a management tool and as a method to combine data sources and algorithms in ways that take advantage of their respective strengths in differing circumstances. The status of tasks that are required to complete this work is also described.
总页数：	1999/10. 9709-9809 pp50
报告类型：	科技报告