Machine learning application powering automation of efficient traffic operation - 博士论文

Aug 5, 2026

Machine learning techniques enable automated, data-driven traffic operations aimed at improving efficiency and congestion management. The dissertation presents a congestion identification and classification workflow using speed-based traffic state estimation and traffic-state clustering, followed by recurrent versus nonrecurrent congestion labeling. It further develops automatic detector mapping metadata generation from high-resolution control inputs, including detector type and phase identification with reported accuracy measures. A computer vision vehicle-tracking study addresses tracking opportunities and challenges, supporting smarter intelligent infrastructure.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages91

Reading Time3 h 49 min

AccessFree to download

Chapter 1. General Introduction

Chapter 2. A Data Driven Method for Congestion Identification and Classification

2.3 Methodology

2.4 Results

Chapter 3. Automatic Generation of Detector Mapping Meta Data Using High Resolution Control Input

3.3 Problem statement

3.6 Results

Chapter 4. Opportunities and Challenges in Vehicle Tracking - A Computer Vision Based Vehicle Tracking System

4.2 (continued)

Machine learning application powering automation of efficient traffic operation by Atousa Zarindast A dissertation submitted to the graduate faculty in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Major: Civil Engineering (Intelligent Infrastructure Engineering) Program of Study Committee: Anuj Sharma, Co-major Professor Christopher Day, Co-major Professor Johnathan Wood Jing Dong Soumik Sarkar The student author, whose presentation of the scholarship herein was approved by the program of study committee, is solely responsible for the content of this dissertation. The Graduate College will ensure this dissertation is globally accessible and will not permit alterations after a degree is conferred Iowa State University Ames, Iowa 2022 Copyright © Atousa Zarindast, 2022. All rights reserved. TABLE OF CONTENTS Page LIST OF FIGURES ....................................................................................................................... iv [LIST OF TABLES ......................................................................................................................... vi](LIST OF TABLES ......................................................................................................................... vi) [ABSTRACT...............................](ABSTRACT...............................)................................................................................................... vii CHAPTER 1. GENERAL INTRODUCTION ................................................................................1 References ................................................................................................................................. 6 CHAPTER 2. A DATA DRIVEN METHOD FOR CONGESTION IDENTIFICATION AND CLASSIFICATION .........................................................................................................................8 2.1. Abstract ........................................................................................................................... 8 2.2. Introduction ..................................................................................................................... 8 2.3. Methodology ................................................................................................................. 12 2.3.1. Study location and description of the data....................................................... 12 2.3.2. Mathematical abstraction of the data ............................................................... 13 2.3.3. Smoothing........................................................................................................ 14 2.3.4. Identifying traffic states based upon speed values .......................................... 15 2.3.5. Congestion detection based on traffic state clustering .................................... 18 2.3.6. Recurrent vs. nonrecurrent congestion classification ...................................... 20 2.4. Results ........................................................................................................................... 21 2.5. Discussion ..................................................................................................................... 23 2.6. Conclusion .................................................................................................................... 25 2.7. References ..................................................................................................................... 26 CHAPTER 3. AUTOMATIC GENERATION OF DETECTOR MAPPING META DATA USING HIGH RESOLUTION CONTROL INPUT......................................................................30 3.1. Abstract ......................................................................................................................... 30 3.2. Introduction ..................................................

FAQs

What is the dissertation’s main goal for traffic operations automation?

To use machine learning to automate efficient traffic operation by identifying and classifying congestion and generating supporting detector/phase metadata.

How does the congestion identification method work?

It estimates traffic states from speed values, applies traffic-state clustering for congestion detection, and then distinguishes recurrent from nonrecurrent congestion.

What does the dissertation contribute regarding detector mapping metadata?

It proposes an approach to automatically generate detector mapping metadata from high-resolution control inputs, including identification of detector types and phase states with accuracy evaluation.

Aug 5, 2026

Upload by Ava Thompson