Safety Analysis of Suburban Arterials in Shanghai, China

Xuesong Wang,Yang Song,Helai Huang
2013-01-01
Abstract:As urbanization accelerates in China, more lands are developed along suburban arterials and more people use these roads to commute. This results in an increase of accesses appearing along the suburban arterial and traffic safety deterioration. This safety problem requires in-depth analysis to understand the relationship between various safety factors and traffic safety. This study examined 161 roadway segments (each between two adjacent signalized intersections) on eight suburban arterials in Shanghai. Geometric design, access features, traffic characteristics and environmental setting for chosen segments were collected. A preliminary inspection of the data revealed the existence of spatial correlation and hierarchical data structure. Therefore, Bayesian Hierarchical models were developed to identify safety factors at both arterial and segment levels. Three hierarchical Negative Binomial (HNB) models were developed for slight crashes, severe crashes and total crashes. Model comparisons between HNB models and Negative Binomial models have affirmed the existence of spatial correlation and hierarchical data structure. Results showed that the density of signal spacing along arterial (DOSP) on arterials has a significant influence on the three types of crashes frequencies. The non-uniform signal spacing has a significant impact on of the occurrence of slight crashes. At the segment-level, higher frequencies of slight crashes, severe crashes and total crashes tend to occur for the segments in inner suburban areas with curves, the segments with a higher density of access points, and the segments with higher percentage of heavy vehicles. The safety impacts of different variables of traffic composite were also studied.
What problem does this paper attempt to address?