Access Type

Open Access Dissertation

Date of Award

January 2023

Degree Type

Dissertation

Degree Name

Ph.D.

Department

Civil and Environmental Engineering

First Advisor

Dr. Stephen Remias

Abstract

Traffic congestion in cities is becoming a more pressing issue as travel demand increases, leading to safety and mobility problems and wasteful use of vehicle energy. Both freeways and signalized corridors are important in creating an efficient transportation network, and congestion on these infrastructure types presents a significant mobility concern in the United States. With advances in sensing technologies for intelligent transportation systems, probe vehicles equipped with GPS-enabled devices such as cell phones and in-vehicle telematics or navigation systems are increasingly being used to gather more information about traffic conditions. While fixed point detectors can be limited by their geographical coverage and sensor density, probe vehicle data offers improved coverage and accuracy. However, proper methods are needed to effectively analyze the data to understand the relationship between congestion factors such as safety and energy, and improve the performance of freeways and signalized intersections. This dissertation aims to extract and utilize new and existing data sources to understand the interrelationship of safety, energy, and freeway and signal performance, with the goal of providing actionable insights for agencies and transportation professionals to mitigate congestion.In this dissertation, we used crowd-sourced probe vehicle trajectory data to examine the relationship between signal performance measures and safety at signalized intersections. While there have been few studies on this topic, our research sought to address spatial unobserved heterogeneity by developing methods at the intersection level. Our findings showed that signal performance is significant in relation to total crash frequency. Next, we used probe vehicle trajectory data to understand the interaction between travel time reliability (TTR) metrics and freeway traffic crashes, while taking into account the heterogeneity and temporal stability of the model. Our research showed that TTR is a crucial factor in reflecting the efficiency of freeway performance and is closely related to recurrent and non-recurrent congestion. Traffic crashes significantly contribute to congestion and undermine mobility in urban areas, so understanding the relationship between TTR and crashes is valuable. Finally, we used onboard diagnostic and probe vehicle trajectory data to develop predictive models for accurately estimating fuel consumption at signalized intersections. Fuel use and emissions are highly variable over time, making fixed time collection points insufficient for capturing fuel use. Our research demonstrated the use of crowd-sourced probe trajectory data to predict intersection-level fuel usage on a signalized arterial, taking into account the non-linear relationship between variables and long-term temporal dependency. Our findings showed that total fuel consumption at an intersection is directly related to traffic operation parameters such as delay. In summary, this dissertation leverages probe vehicle trajectory data to build models that quantify the effects of key factors contributing to congestion: safety, energy, freeway performance, and signalized intersection performance. The goal is to provide transportation agencies with insights to help mitigate congestion.

Download

Share

COinS