Effectiveness of Consistency Measures in Crash Prediction Models for Two-Lane Highways in Palestine

Abstract

Crash prediction models (CPMs) are important tools to predict the safety performance of highways based on their operational and geometric characteristics. Most of these models use geometric design characteristics to predict the number of crashes. However, few studies have identified the effectiveness of geometric design consistency measures in explaining crash occurrence probability and crash prediction. Crash prediction models that incorporate design consistency measures for rural two-lane, two-way roadways in the West Bank of Palestine have been developed. This study compares the effectiveness of the resulting models with those which rely on geometric design characteristics only. Approximately 118�km length of two-lane rural highways in the West Bank was selected covering a variety of highway classifications, locations, and characteristics. Different design consistency measures were considered in the developed CPM. The generalized linear regression modeling approach was used, and different mathematical forms were tested. The research presents a quantitative comparison between CPMs that explicitly consider design consistency with the Highway Safety Manual models, which use geometric design characteristics. The study proposes an approach to identify geometrically inconsistent locations using the safety-consistency factor. The results indicate that geometric consistency measures could provide more reliable crash prediction models, thus predicting collision potential more accurately. � 2018, King Fahd University of Petroleum & Minerals.