A Crash Severity-Based Black Spot Identification Model

Abstract

The objective of this research is to develop a new black spot model incorporating the severity of crashes with high consistency. Crash history from 2005 to 2009 of the Ipswich Motorway, in Australia, was used to identify black spots by using simple ranking and empirical Bayes approach. The results of these analyses were then compared via two consistency tests. These tests revealed that for the simple ranking method and the empirical Bayes method, the crash type analysis was more consistent than the crash frequency analysis and the collective risk analysis. Further research into the validity of the proposed indicator would be recommended due to its incorporation of the crash severity and the high consistency it demonstrated in this study, both of which are of great benefit from a practical point of view.

Development of New Performance Measures Based on Data Mining Weights for Hotspot Identification

“…Previous research tried to reflect severity indicators in the hotspot identification process in various ways ( 14, 15, 17, 32 ). In this study, severity classification models based on the data mining method were developed.…”

“…These two studies show that it is difficult to prove statistical significance because severity data have an uncertainty distribution. Thus, in most studies, step-by-step joint modeling was performed for accurate modeling and to capture the uncertainty distribution of severity data ( 17 ). These studies performed additional severity analysis separately from frequency-based hotspot identification.…”

Collective Risk Ranking of Highway Segments on the Basis of Severity-Weighted Crash Rates

“…Efforts were also made to introduce a combined index used for ranking potential hotspots, providing a reference to impose safety countermeasures. Da Costa et al [7] incorporated crash severity in a collective risk ranking of road segments in Australia. e severity-weighted frequencies per crash type (based on crash severity costs in Australia) were adjusted for a potential RTM bias via the EB method.…”

“…where P itj is the probability of crash severity j occurring on segment i in year t, C j is the cost for severity level j, and M it is the total number of crash predictions or observations in segment i and year t. When S P is calculated, P is taken directly from the classifier probability predictions, while for S O , P is simply taken as the ratio m j /N O . Finally, N E is modified to estimate the collective risk (CR) by predicting the weighted crash rate S F for future year t * (i.e., the year 2010 or 2011) [7].…”

Evaluation and Improvement of Carrying Capacity of a Traffic System

“…At the same time, there will be new buildings in research region which means the traffic system will face new challenges (shown in Figure 3). Therefore, in this research, if the traffic problems can be solved, the recommendations and measures used in this region can also be referenced to other traffic systems [5][6][7].…”