Semester of Graduation
Summer 2018
Degree
Master of Science in Civil Engineering (MSCE)
Department
Civil and Environmental Engineering
Document Type
Thesis
Abstract
Traffic accidents in the U.S. have declined over the last two decades but continue to cost the country billions of U.S. dollars each year. Intersection accidents alone account for 23% of the 32,675 deaths that occurred by motor vehicle crashes in 2014. In light of this, the first part of this study sought to identify driver behavior and secondary tasks that influenced the occurrence of a crashes at controlled, uncontrolled and non-intersections with multiple logistic regression. The second part of the study used decision tree which applied machine learning algorithms to identify significant factors that influenced the occurrence of crashes at intersections and nonintersections. Multiple logistic regression results revealed that driver behavior and secondary tasks that affected crash events at controlled intersections were stop sign violations, signal violations, improper turns, exceeding speed limit and performing visual-manual tasks. Also, the driver behavior and secondary tasks that affected crash events at uncontrolled intersections were improper turn, visual manual tasks and exceeding the speed limit. At non-intersections visual manual tasks and exceeding the speed limit were selected as significant contributors to crashes. The decision tree for intersections determined that driver behavior, locality, v1 lane occupied, alignment and through travel lanes were the primary splitters in the tree. If the driver was distracted or exhibited no driving behavior before a crash and was driving on a dedicated left turn lane, 89% of the time, the crash occurred at an intersection. Also, all driving behavior, apart from distracted driving, took place before intersection-related crashes 88% of the time emphasizing that intersections are risky locations. The findings of this study are very promising and demonstrate the importance of obeying traffic rules and regulations. These findings could also serve as a guide for departments of transportation in the various states to choose locations for safety improvement programs.
Date
5-15-2018
Recommended Citation
Ashley, Grace Naa, "Effect of Driver, Vehicle, and Road Related Factors on Crashes at Intersections and Non-Intersections: A Naturalistic Driving Study" (2018). LSU Master's Theses. 4724.
https://repository.lsu.edu/gradschool_theses/4724
Committee Chair
Ishak, Sherif
DOI
10.31390/gradschool_theses.4724