Graduation Semester and Year
Summer 2024
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Civil Engineering
Department
Civil Engineering
First Advisor
Pengfei(Taylor) Li
Second Advisor
Stephen P. Mattingly
Third Advisor
Kate Hyun
Fourth Advisor
Xianfeng (Terry) Yang
Abstract
This article-based dissertation consists of four complete manuscripts related to using proposed advanced network and graph theoretical approaches to optimize traffic control. Manuscript One addresses the integration of Connected Automated Vehicles (CAVs) and human-driven vehicles in heterogeneous traffic networks. Traditional models assume uniform driving behaviors, which are now inadequate. To address this, an integer programming formulation for a vehicle-oriented traffic dynamics model in space-time networks is introduced. This model distinguishes between individual vehicles and their unique driving behaviors, capturing diverse attributes of heterogeneous traffic flow. The manuscript demonstrates how the model can estimate the propagation of backward shockwaves for scheduled CAVs based on each queuing vehicle's delay, free-flow travel time, and travel speed in the queue. This framework lays the foundation for controlling queue evolution and traffic assignment in large-scale heterogeneous traffic networks.
Manuscript Two focuses on developing advanced traffic control strategies leveraging emerging traffic data from mobile computing. It introduces a new multi-intersection phase (MI-phase) to facilitate safe vehicle movements across tightly connected intersections. This model treats interconnected intersections as a single "super intersection," where vehicles move according to planned paths, minimizing delays. A linear integer programming formulation for optimizing vehicle space-time trajectories and traffic control is presented. A scalable optimization framework, "Lagrangian decomposition with subproblem approximation," incorporates a dynamic network loading-based lower bound estimator (DNL-LBE), managing complex controlled dynamic network loading processes efficiently using Lagrangian multipliers and Dynamic Programming. The approach enhances traffic control coordination and demonstrates efficiency through advanced computing techniques.
Manuscript Three presents an adaptive traffic control strategy for urban signalized interchanges, utilizing traditional sensors and CAV technologies. This strategy targets interchanges controlled by a single traffic signal controller where vehicles must cross multiple stop lines, such as diamond interchanges (DIs), diverging diamond interchanges (DDIs), and single-point urban interchanges (SPUIs). A new adaptive traffic control framework based on a phase-time network is proposed, dynamically fine-tuning control splits and optimizing the phasing sequence based on vehicle arrival counts from infrastructure sensors and turning movement ratios from CAV technologies. The optimization problem is formulated as a mixed-integer linear programming (MILP) problem, validated through offline examples, and evaluated using an online search algorithm within a microscopic traffic simulation environment. Results demonstrate the efficacy of the MILP formulation and algorithm, showing promise for real-world implementations.
Manuscript Four presents my latest research, introducing an advanced traffic signal optimization model based on the cell transmission model (CTM) and enhanced by an innovative offset-enabled distributed phase-time network. This model integrates CTM with signal control variables to accurately capture traffic dynamics. The optimization framework targets minimizing network delays and maximizing bandwidth efficiency by optimizing offsets and green-band width within the phase-time network. The model's efficacy is validated through experiments under congested traffic conditions, utilizing various traffic signal optimization strategies. Additionally, the paper proposes an optimization model employing Lagrangian relaxation, which decomposes the original problem into signalized and non-signalized subproblems, thereby simplifying computational complexity and enhancing the model's scalability and applicability.
In conclusion, these manuscripts collectively introduce innovative models and frameworks for optimizing traffic control in heterogeneous and congested urban environments, leveraging both traditional and emerging technologies. The proposed approaches demonstrate improved efficiency and scalability, providing robust solutions for real-world traffic management challenges.
Keywords
Traffic control, Multi-intersection, Cell transmission model, Traffic coordination, Band maximization, Offset optimization, Phase time network, Problem approximation, Lagrangian decomposition
Disciplines
Civil Engineering | Transportation Engineering
License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Recommended Citation
Wang, Peirong, "ADVANCED NETWORK AND GRAPH THEORETICAL APPROACHES TO URBAN TRAFFIC CONTROL AND OPTIMIZATION" (2024). Civil Engineering Dissertations. 289.
https://mavmatrix.uta.edu/civilengineering_dissertations/289
Comments
I want to express my deepest gratitude to my committee chair, Dr. Pengfei (Taylor) Li, for his unwavering dedication and commitment to assisting me in my academic endeavors. Dr. Li's genuine devotion to the transportation profession and research is exemplary and admirable. I am immensely grateful for his recognition of my potential when I was just out of my bachelor's degree with little knowledge in transportation engineering. His belief in me, demonstrated through his encouragement to pursue challenging projects and his trust in my abilities, has been a driving force throughout the seven years we have worked together, and his patience and exceptional guidance have been instrumental in my growth.
Secondly, I extend my sincere gratitude to Dr. Stephen P. Mattingly and Dr. Kate Hyun for organizing the transportation graduate student research meetings. These meetings were not just a platform for me to contemplate the transportation problems I wanted to solve and present my ideas, but they were also a catalyst in my academic journey. The peer feedback I received, which I deeply value, was instrumental in helping me complete the third manuscript. I would also like to thank Dr. Xianfeng (Terry) Yang for the transformative opportunity to work with him on the Utah Department of Transportation (UDOT) project. This experience allowed me to delve into cutting-edge, connected vehicle technology and research, significantly enhancing my academic growth.
I am very grateful to my family—my parents and in-laws. Thank you for always being there for me on this rough and bumpy road. I believe I have made all of you proud. Special thanks to my wife, Shuaiqi Hu, for her support and encouragement during the challenging times. Her patience and unconditional support have been my pillars of strength, and I would not have made it to this point without her.
Additionally, my deepest appreciation goes to my lab mates, including but not limited to Dr. Farzan Rahman Chowdhury, Swastik Khadka, and Sijan Shrestha. I also thank my friends Jiayong Lin, Zhaoyuan Sun, Siyu Wang, Rongrong Tao, for their endless encouragement and company.
Peirong (Slade) Wang
The University of Texas at Arlington, July 2024