Modeling drivers’ naturalistic driving behavior on rural two-lane curves

Thumbnail Image
Date
2015-01-01
Authors
Wang, Bo
Major Professor
Advisor
Shauna Hallmark
Jing Dong
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Authors
Research Projects
Organizational Units
Journal Issue
Is Version Of
Versions
Series
Department
Civil, Construction, and Environmental Engineering
Abstract

This dissertation examined drivers’ naturalistic driving behavior on rural two-lane curves using the Strategic Highway Research Program 2 Naturalistic Driving Study data. It is a state-of-the-art naturalistic driving study that collected more than 3,000 drivers’ daily driving behavior over two years in the U.S. The major data sources were vehicle network, lane tracking system, front and rear radar, driver demographics, driver surveys, vehicle characteristics, and video cameras. This dissertation has three objectives: 1) examine the contributing factors to crashes and near-crashes on rural two-lane curves; 2) understand drivers’ normal driving behavior on rural two lane curves; 3) evaluate how drivers continuously interact with curve geometries using functional data analysis.

The first study analyzed the crashes and near-crashes on rural two-lane curves using logistic regression model. The model was used to predict the binary event outcomes using a number of explanatory variables, including driver behavior variables, curve characteristics, and traffic environments. The odds ratio of getting involved in safety critical events was calculated for each contributing factor. Furthermore, the second study focused on the analysis of drivers’ normal curve negotiation behavior on rural two-lane curves. Significant relationships were found between curve radius, lateral acceleration, and vehicle speeds. A linear mixed model was used to predict mean speeds based on curve geometry and driver factors. The third analysis applied functional data analysis method to analyze the time series speed data on four example curves. Functional data analysis was found to be a useful method to analyze the time series observations and understand driver’s behavior from naturalistic driving study.

Overall, this dissertation is one of the first studies to investigate drivers’ curve negotiation behavior using naturalistic driving study data, and greatly enhanced our understanding about the role of driver behavior in curve negotiation process. This dissertation had many important implications for curve geometry design, policy making, and advanced vehicle safety system. This dissertation also discussed the opportunities and challenges of analyzing the Strategic Highway Research Program 2 Naturalistic Driving Study data, and the implications for future research.

Comments
Description
Keywords
Citation
Source
Copyright
Thu Jan 01 00:00:00 UTC 2015