Degree
Doctor of Philosophy (PhD)
Department
The Department of Geography and Anthropology
Document Type
Dissertation
Abstract
The safety of highway-railroad grade crossings (HRGC) is still an issue in the United States of America (USA). The grade crossing is where a railroad crosses a road at the same level without any over or underpass. To improve the safety of crossings, the crossings’ condition should be explored from several aspects such as engineering design (speed limit, warning signs, etc.), road condition (number of lanes, surface markings, etc.), rail design (the type of track, ballast, etc.), temporal variables (weather, visibility, time of day, lightning, etc.), social variables (population, race, etc.), and last but not least, spatial variables (the type of land use, distance to nearby intersections, distance to nearby crossings, distance to nearby specific land uses including emergency medical services (EMS), etc.).
It has always been a hot topic for the rail traffic safety projects to eliminate the number of grade crossings by closing redundant crossings in an area. The closure of a few HRGC in a specific geographical area is known as consolidation programs. Several studies have been done so far to define whether a crossing in a neighborhood needs safety improvements, level separations, or closure programs. In this research, I aim to investigate the applicability of spatial analysis, machine learning, and text mining methods to study the safety of HRGC. To do so, two datasets of HRGC data and HRGC crash data were downloaded and used from the Federal Railroad Administration (FRA). Since HRGC is a nationwide problem, first I considered the case study to the extent of the USA. Then I narrowed down to the state of Louisiana to create a more reliable consolidation program. Finally, focusing on the spatial variables that may influence closure programs, I developed a consolidation model for a smaller area, East Baton Rouge Parish (EBRP).
Hopefully, this research could assist rail safety personnel by making better decisions on highway-rail grade crossing closures and thus would benefit residents of a city experiencing a safer area to reside. Furthermore, it is possible that results of this research could be extended to other states, where they face similar problems with the safety of HRGC.
Date
10-15-2020
Recommended Citation
Soleimani, Samira, "Using Spatial Analysis and Machine Learning Techniques to Develop a Comprehensive Highway-Rail Grade Crossing Consolidation Model" (2020). LSU Doctoral Dissertations. 5369.
https://repository.lsu.edu/gradschool_dissertations/5369
Committee Chair
Leitner, Michael
DOI
10.31390/gradschool_dissertations.5369
Included in
Data Science Commons, Geographic Information Sciences Commons, Spatial Science Commons, Transportation Engineering Commons