1 |
Evaluating the Safety Effects of Signal ImprovementsDowell, Ashley Lynn 23 May 2013 (has links) (PDF)
As a result of high crash frequencies on roadways, transportation safety has become a high priority for the United States Department of Transportation and the Utah Department of Transportation. A large percentage of fatal and injury crashes on roadways occur at intersections and traffic signals have been implemented to reduce these severe crashes. There is a need to evaluate the effectiveness of the traffic signal improvements through the development of Crash Modification Factors (CMFs). Recent research has shown that traditional safety evaluation methods have been inadequate in developing CMFs. In recent years, Bayesian statistical methods have been utilized in traffic safety studies to more accurately analyze the effectiveness of safety improvements. The hierarchical Bayesian method is an advanced statistical technique that has the capability to account for the shortcomings of traditional methods and to more fully reflect the effectiveness of safety improvements. This report uses a hierarchical Bayesian model to analyze the effectiveness of new traffic signal installations and modified traffic signals. CMFs were developed for multiple scenarios for both new and modified traffic signals. A benefit-to-cost (B/C) analysis was also performed for each improvement to determine how long it would take to recover the cost of installation. The results showed that there was an increase in overall crashes for both new signal installations and modifications to existing signals. The severe crash analysis revealed that there was an increase in non-severe crashes and a reduction in severe crashes; the improvements are effectively reducing severe crashes and improving safety at intersections. The B/C analyses indicate that there is a safety benefit to both improvements and that new signal installation costs can be recovered in approximately 5 years while the installation of a left-turn signal modification can be recovered in approximately 9 weeks.
|
Page generated in 0.1121 seconds