The effects of selected crash parameters and traffic volume on crash frequency (the case of Bahirdar city, Ethiopia)
No Thumbnail Available
Date
2018-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
AAU
Abstract
The aim of the study is to develop mathematical model to understand the impact of selected crash
parameters on crash frequency in in high and low traffic volume in both straight section and
junction. Comparing the effect of straight section and junction with selected crash parameters on
crash frequency is also addressed. Four models are developed, these are crash condition at high
traffic volume in straight section, low traffic volume in straight section, high traffic volume in
junction and low traffic volume in junction.
The method used in this research is loglinear analysis using SPSS software. Seven days’ traffic
data was collected from major junction and straight section in Bahirdar. The purpose of traffic data
collection was to categorize hours of the day into high traffic volume hours and low traffic volume
hours. This classification helps to know the relationship between traffic crash and traffic volume
at straight section and junction. Five-year traffic crash was also collected to conduct the analysis
and modelling in SPSS software by loglinear modelling. Variable selection and coding were also
addressed in methodology part.
The result depicted that in high traffic volume straight section Monday has the highest probability
of crash occurrence compared to base case, Sunday. In low traffic volume junction Wednesday
has the highest probability of crash occurrence compared to Saturday. in high traffic volume
straight sections 18-29-year age group 3.5 times higher probability of crash occurrence compared
to base case (>45 year). 30-44 year drivers have also 2.1 times probability of crash occurrence
compared to >45 year drivers. In low traffic volume straight sections both 18-29 and 30-44-yearold
drivers
have
increased
the
probability
of
crash
occurrence.
In
low
traffic
volume
junction
both
18-29
and
30-44-year-old
drivers
have
increased highly with 47.9 and 38.1 times the base case
(>45 year). In low traffic volume straight section high probability of crash occurrence is observed
in < 5 years driving experience compared to the base case (>15 years). In low traffic volume
junction <5 years driving experience has also high probability of crash occurrence compared to
the base case. In high traffic volume straight section and low traffic volume junction road with
median has 90 % less probability of crash occurrence compared to road without median. In low
traffic volume straight section road with median has 70 % less probability of crash occurrence
compared to road without median. In high traffic volume priority junctions have 3.5 times higher probability of crash occurrence compared to round about. And 60% less probability of crash
occurrence in signalized junction compared to roundabout.
Description
Keywords
log linear modelling, traffic crash, traffic volume, junction