Network Resilience Analysis and Traffic Control Strategies Under Infrastructure Failures and Disaster Response

This dissertation investigates strategies to enhance the robustness and efficiency of transportation networks under severe disruptions, particularly infrastructure failures and disaster scenarios. It presents a comprehensive framework for monitoring and managing network resilience, and explores the potential of electric vehicles (EVs) to facilitate more effective traffic evacuation.
Firstly, the dissertation develops a resilience assessment methodology to evaluate transportation network performance following infrastructure disruptions, exemplified by the closure of the Hernando de Soto Bridge on Interstate 40. Integrating the Resilience Triangle framework with advanced statistical methods, this study underscores the critical influence of large-scale queue distributions and the heavy-tailed characteristics of extreme congestion events.

Secondly, the research introduces a percolation-based evacuation strategy designed to prevent rapid gridlock in scenarios of urgent evacuation. Through microscopic simulations, it demonstrates how strategically monitoring and regulating the inflow of vehicles can mitigate widespread spillback congestion and accelerate evacuation efficiency.

Lastly, this dissertation evaluates how the unique characteristics of electric vehicles can enhance evacuation processes and overall network resilience. Empirical analyses highlight EVs' superior agility, including accelerated lane-changing capabilities and rapid acceleration compared to traditional gasoline-powered vehicles. Furthermore, a novel model predictive control (MPC) framework integrated with the MOBIL lane-changing model demonstrates how EV agility facilitates efficient lane-changing behaviors, enabling evacuees to utilize available traffic gaps effectively. The study also explores EVs' implications for autonomous vehicle (AV) performance during evacuations, concluding from experimental data that EVs exhibit lower delays and greater precision in executing AV trajectories. Real-time analyses further illustrate that the operational advantages of EVs can reduce shockwave propagation and enhance traffic throughput during peak demand or emergency conditions.

Collectively, the findings contribute novel insights into designing transportation networks capable of withstanding routine stresses and dynamically responding to significant disruptions. The outcomes of this research provide valuable guidance for transportation engineers, planners, and policymakers tasked with developing resilient, efficient, and future-oriented infrastructure systems.