REVOLUTIONIZING URBAN TRAFFIC MANAGEMENT USING DEEP REINFORCEMENT LEARNING

Author

Tanvi Reddy Dasaripally, Southern Illinois University CarbondaleFollow

Date of Award

5-1-2024

Degree Name

Master of Science

Department

Computer Science

First Advisor

Hexmoor, Henry

Abstract

This thesis explores the integration of Deep Reinforcement Learning (DRL) in urban traffic management, examining how these technologies can be adapted and their effectiveness in the context of growing intelligent city infrastructures. By thoroughly examining more than 30 advanced research studies, this work highlights the creative use of DRL and other related technologies, such as vehicle-to-everything (V2X) communications, to improve traffic flow, alleviate congestion, and facilitate the smooth operation of Connected and Automated Vehicles (CAVs) in urban settings. This thorough review emphasizes the potential of DRL to surpass traditional traffic management systems, showcasing its ability to adapt to complex and dynamic traffic scenarios.Nevertheless, this thesis delves deeper into the analysis to pinpoint notable gaps and challenges that continue to exist in the current research landscape. These tasks involve integrating DRL-based systems with existing traffic management infrastructures, developing DRL models that can be applied to various urban environments, moving from simulated studies to real-world implementation, and considering the ethical considerations of using AI in public areas. Tackling these challenges is essential for unlocking the full potential of DRL in urban traffic management.The proposed future directions aim to address the identified research gaps and envision a future where traffic management systems are more advanced, streamlined, and fair. This thesis asserts that by adopting these cutting-edge solutions, urban traffic management can make substantial progress, resulting in enhanced mobility, safety, and quality of life in urban centers across the globe. The effective execution of these proposals has the potential to revolutionize urban traffic management, creating a system that is more adaptable, environmentally friendly, and accessible. This will set the stage for the development of intelligent cities in the years to come.