Doctoral Dissertation Research: An Integrated Agent-Based Microsimulation Model for Hurricane Evacuation in New Orleans

飓风造成的城市地区大规模疏散是一个关键问题，需要进行广泛的基础和应用研究。 以前的研究使用微观模拟方法来模拟单个车辆和驾驶员的行为，但主要限于在一个小的研究区域的模拟。 很少有研究考虑在疏散策略中使用脆弱性。 有必要开发一种更好的方法来整合各种信息（如脆弱性），以更好地模拟疏散，以便我们能够及时有效地疏散人员。 这个博士论文研究改进项目将开发一个集成的微观模拟模型来模拟新奥尔良的飓风疏散。 本计画的第一个目标是整合两个以代理人为基础的微观模拟模型，建立一个两层的区域性灾害疏散模型。 第一个模型将把每辆车作为一个代理人，以模拟区域道路网络的交通运动的高速公路，而第二个模型将使用每个人口普查块质心作为代理人，以模拟区域道路网络的交通的本地部分。 第二个项目目标是评估三种不同区域疏散策略的总体疏散时间和其他指标的有效性，包括同时疏散策略，基于研究区域不同社区疏散脆弱性的分阶段疏散策略，以及基于不同社会背景的分阶段疏散策略该研究项目将有助于动态空间建模、可视化以及灾害和脆弱性科学领域。 将脆弱性纳入两种类型的基于代理的模型将提供更真实和准确的图片如何疏散行为影响不同的疏散策略的有效性，从而提供有用的指导政策制定者。 虽然所提出的综合基于代理的模型将在新奥尔良飓风疏散的基础上，开发的方法和从这项研究中获得的知识将适用于其他地区和其他类型的灾害。 此外，在本研究中开发的计算机模块将被广泛分布，以有益于其他基于代理的交通微观仿真模型。 作为博士论文研究改进奖，该奖项还将提供支持，使有前途的学生建立一个强大的独立的研究生涯。

Mass evacuation of urban areas due to hurricanes is a critical problem that requires extensive basic and applied research. Previous research used a microsimulation approach to simulate individual vehicle and driver behavior but was limited mostly to simulation in a small study area. Very few studies have considered the use of vulnerabilities in the evacuation strategies. There is a need to develop a better approach to integrate the various information (such as vulnerabilities) to better simulate evacuation so that we can evacuate people timely and efficiently. This Doctoral Dissertation Research Improvement project will develop an integrated microsimulation model to simulate hurricane evacuation in New Orleans. The first objective of this project will be to create a two-level regional disaster evacuation model by integrating two agent-based microsimulation models. The first model will treat each vehicle as an agent to simulate regional road network traffic movements on highways, while the second model will use each census block centroid as an agent to simulate the local part of the regional road network traffic. The second project objective will be to evaluate the effectiveness in terms of overall evacuation time and other metrics of three different regional evacuation strategies, including simultaneous evacuation strategy, a staged evacuation strategy based on different community-evacuation vulnerabilities over the study area, and a staged evacuation strategy based on different social-biophysical vulnerabilities.This research project will contribute to the fields of dynamic spatial modeling, visualization, and disaster and vulnerability science. The incorporation of vulnerabilities into two types of agent-based models will provide more realistic and accurate pictures of how evacuation behavior affects the effectiveness of different evacuation strategies, thus providing useful guidance to policy makers. Although the proposed integrated agent-based model will be based on hurricane evacuation in New Orleans, the methods developed and the knowledge gained from this research will be applicable for other regions and other kinds of disasters. Furthermore, the computer modules developed in this research will be widely distributed to benefit other agent-based traffic microsimulation models. As a Doctoral Dissertation Research Improvement award, this award also will provide support to enable a promising student to establish a strong independent research career.