Use of Dynamic Incident Progression Curve for Classifying Secondary Accidents

Abstract

Secondary accident statistics can be useful in studying the impact of traffic incident management strategies. A methodology is presented for classifying secondary accidents by using data fusion of police accident database with intranet incident reports. One current method for classifying secondary accidents uses a static threshold that represents the spatial and temporal region of influence of the primary accident, such as 2 miles and 1 hour. In other words, an accident is considered secondary if it occurs upstream from the primary accident and is within the duration and the queue of the primary accident. However the use of the static threshold could result in both false positives and negatives since accident queues are constantly varying. The methodology presented seeks to improve upon this existing method by making the threshold dynamic. In fact an incident progression curve is used to mark the end of the queue throughout the entire incident. Four steps in the development of incident progression curves are described. Step one is the processing of intranet incident reports. Step two is the filling in of incomplete incident reports. Step three is the non-linear regression of incident progression curves. Finally, step four is the merging of individual incident progression curves into one master curve. 5514 accidents from Missouri freeways are analyzed and the results show that the use of dynamic versus static thresholds can differ by more than 30%.