Dynamic Mode Control for Transit Services in Disaster Evacuations Using a Hawkes-Based Jump-Diffusion Process
Guardado en:
| Publicado en: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Autor principal: | |
| Publicado: |
ProQuest Dissertations & Theses
|
| Materias: | |
| Acceso en línea: | Citation/Abstract Full Text - PDF |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Resumen: | This dissertation introduces a new strategic decision architecture for dynamic transit mode and vehicle size switching during mass evacuations, a concept previously unexplored in the evacuation literature. In this approach, we integrate transit operation cost functions with evacuation-specific metrics such as congestion-related delays, unmet demand, and time penalties, to form an evacuation-metrics-in-the-loop optimization structure. Given the inherent constraints of community resources and government subsidies, this framework addresses the practical challenges of efficiently allocating limited transportation agency fleets and vehicle capacities to meet stochastic evacuation demand. To achieve this, we propose an Immediate Cost-Savings Strategy, where dynamic control is guided by near-term evacuation cost reductions, while disregarding switching costs. Furthermore, to model the problem more realistically and address uncertainty with abrupt surges in stochastic evacuation demand, we incorporate the Jump-Diffusion Process (JDP) into evacuation literature for the first time. We then extend the model by integrating a Hawkes process with a jump-diffusion framework, forming a stochastic process where jumps are correlated and influence future jumps. Given the scarcity of real-world evacuation demand density data, we demonstrate the potential of JDP through simulation experiments to effectively capture dynamic demand fluctuations triggered by disaster announcements, aftershocks, or infrastructure disruptions. Simulations under static and stochastic demand scenarios identify critical thresholds when switching between transportation services (fixed and flexible) or vehicle sizes (low-capacity and high-capacity) optimizes cost-effectiveness and evacuation performance. Overall, this study offers a new perspective on evacuation planning by integrating dynamic mode-switching decisions with stochastic demand modeling within the evacuation context. By identifying the exact demand thresholds that necessitate these adjustments, our approach captures the real-world challenges of optimizing limited fleet resources under the time-critical pressures of evacuation scenarios. The remainder of this dissertation is structured with a progression that builds from foundational concepts to advanced modeling and policy evaluation. It begins in Chapter 1 with a thorough review of existing literature on simulation – and optimization-based studies in evacuation planning, through which critical research gaps are identified by highlighting the study’s motivation. This review flows into an outline of the main contributions and key elements of the proposed framework. Building on this, Chapter 2 introduces the proposed evacuation policy and its overarching decision architecture by delving into the details of each component and illustrating their interconnections to reveal how they collectively support the decision-making framework. The analysis then transitions into Chapter 3, where the static model configuration is presented. Here, the target network is defined, and baseline demand is established using static assumptions for the identification of key threshold points for switching between service sizes and transit modes. These insights are critical for shaping Chapter 4, where the dynamic model takes shape with the introduction of the JDP. This chapter highlights the JDP’s relevance and advantages in capturing dynamic evacuation demand by presenting a detailed rationale for parameter selection and results from computation experiments and sensitivity analysis. By pinpointing the most influential JDP parameters, the chapter integrates dynamic simulations with time-varying demand to demonstrate the proposed policy’s effectiveness and quantifying its cost-saving potential. Finally, Chapter 5 ties the dissertation together by discussing the study’s key limitations and proposing potential research directions, before concluding with final remarks that summarize the work’s contributions and broader implications. |
|---|---|
| ISBN: | 9798291598665 |
| Fuente: | ProQuest Dissertations & Theses Global |