【作者】 张钊；

【导师】 晏启鹏；

【作者基本信息】 西南交通大学 ， 交通工程， 2013， 博士

【摘要】 过去的几十年以来,全球气候不断变化,气温不断升高,冰川融化,海平面上升,近十年以来海啸、飓风(我国称之为台风)频繁发生,袭击全球各地沿海区域,造成人员丧生和财产损失。诸如袭击中国上海、浙江、福建等地的台风,袭击美国密西西比州、路易斯安那州、德克萨斯州的飓风katrina,Rita、Ike、Gustav等。全球气候变暖、海啸飓风爆发频率增加的同时,城市化进程不断进行：海岸线人口不断增加、集聚,随之带来城市发展的一个现象：城市群(mega region)的形成。近五十年来,全球人口不断增长,在一些区域,原本相距较远的单独城市,城市人口和城市规模不断增长,多个城市最终连接成人口分布在较广区域,延续几百公里的城市组团,其被称为城市群。目前,我国有三个典型城市群：长三角城市群、珠三角城市群和环渤海城市群；美国有两个典型的城市群：波士顿-纽约-费城-巴尔的摩-华盛顿哥伦比亚特区城市群、南加州圣地亚哥-洛杉矶城市群和多个潜在的城市群。在世界范围内,类似的城市群在印度、巴西、日本、南非、欧洲等世界各地迅速扩张和发展。伴随着全球气候变暖、灾难性事件的频发和城市群的快速发展。突发性气候灾难威胁城市群数以百万的人口的可能性在不断增加。灾难性事件下应急管理方法之一是采用大区域交通疏散,将危险区域的人员疏散到安全区域,即应急交通疏散。然而,包括数百万车辆,数千万人口的城市群应急交通疏散尚未被研究者纳入实际研究领域,本研究将就城市群应急交通疏散的模型构建、分析方法作一个初步的探索。本文提出城市群应急交通疏散模型参数要求和四阶段建模步骤,构建一个城市群应急交通疏散微观仿真模型,该仿真模型将被用于城市群的应急交通疏散研究。研究中使用一个理论上的城市群,即墨西哥湾城市群(路易斯安娜州新奥尔良大都市圈延伸至德克萨斯州休斯敦大都市圈)作为研究区域,尽管该城市群的规模尚不够巨大,但其仍涵盖了1200万人口,且在未来时间内,人口还会继续增加；选择该区域的另—原因在于研究者所在课题组已经构建过新奥尔良区域的交通仿真模型,因而,在现有研究基础上,可以将模型区域向西扩展,进而构建一个涵盖6个大都市圈的城市群应急交通疏散模型。该模型融合地理信息系统,疏散需求预测模型和交通仿真分析平台(The Trans-portation Analysis and Simulation System, TRANSIMS)构建,能够整合利用现有的人口数据和地理信息数据来构建人口和网络信息,模型弹性较好,能够适应未来人口增长和道路网络的变化。研究中,构建虚拟飓风情境,进行模型实验再现疏散演化过程,人口变化、道路网络、疏散者行为以及区域管理策略都将被测试,以研究上述变量和交通系统表现之间的关系。从中得出在不同情形下的关键通道速度时空分布图,流量时空分布图,网络瓶颈位置,网络清空时间,宏观网络交通有效性指标等。进一步分析表明：疏散交通需求、交通管理策略对于交通表现指标有决定性的作用,通过曲线拟合的方法,构建了网络交通表现函数,以预测在不同的疏散需求和管理策略组合下,城市群道路网络的综合表现指标值,以及在需求和管理策略发生变化的情况下,城市群道路网络的疏散交通状况会发生何种改变,研究了网络清空时间的数学分布和分区域分时段疏散策略制定的数学方法,基于此提出了城市群应急交通疏散策略制定方法和步骤,为应急交通管理者和决策者提供信息支撑,做出合理决策。由于飓风的多发性和典型性,人们在飓风下的疏散行为和其他可预知灾难性事件下的行为类似,以上研究成果不仅有助于人们更加深入理解城市群应急交通疏散现象,而且可以将其应用于不同区域,人口、道路网络结构城市群,包括飓风在类的可预知灾难性事件下的应急交通规划、管理和分析。更多还原

【Abstract】 Over the past decade, there has been a growing census of long range climatological forecasters that the earth is experiencing significant changes in its climate. These changes, particularly those related to ocean warming and sea-level rise threaten coastal regions throughout the world. More active and severe weather patterns thought to be linked to the-se conditions appear to be increasing both the frequency and severity of tropical storm sys-tems. Adding to the level of threat over this same period has been a parallel explosion of population growth along coastal regions worldwide. These increases have given rise to a new phenomenon or urban development, known as the mega region.Mega regions are continuously populated areas that have grown together over time from distinctly separate individual cities or populated areas to form continuously densely popu-lated areas that may span over many hundreds of miles. In the United States, there are two obvious mega regions, including the Boston-New York-Philadelphia-Baltimore-Washington DC area and the San Diego-Los Angeles region of Southern California with the potential to have several more in Florida, the Midwest-Great Lakes, and the Gulf Coast areas within the next25to50years. These areas are dwarfed, however, by mega regions in other areas of the world. The Hong Kong, Shenhzen, and Guangzhou Region of China, that includes about120million people is considered the world’s first mega region. Similar trends are occurring elsewhere in China as well as in India, Brazil, Japan, and West Africa.When the trends of climate change and population growth are combined, it is inevitable that there will also be a significant increase in the number of catastrophic disasters that can threaten millions of people. With this realization also comes the need to plan and prepare for such occurrences. One of the basic tools of emergency management is the use of re-gional mass evacuations. However, the idea of conducting a mega region evacuation that could include tens of millions of people has never been realistically considered. This re-search seeks to take the initial steps toward investigating the mega region evacuation issue.In the research, a traffic simulation model was developed to analyze traffic conditions associated with a mega region mass emergency evacuation. Similar to the growth process of actual mega regions, the mega region test case used in the research was developed for a theoretical mega region along the US Gulf Coast from New Orleans to Houston. While this area cannot yet be considered a true mega region, it does encompass a population of about12million people with the potential for considerable growth in the future. It was also se-lected because two regional evacuation traffic simulation models have already been devel- oped for New Orleans and Houston. Thus, the test mega region was formed by "con-necting" those two separate models with those of the Baton Rouge, Lafayette, Lake Charles, and Beaumont regions in between.The model was constructed using the TRANSIMS mesoscopic activity-based transpor-tation modeling system. TRANSIMS was thought to be well-suited to the research plan because it permits the development of the simulation because it is able to utilize existing census and geographical data to build populations and networks. These can then be modi-fied to represent future growth and development with the region. In the research, a range of population, road network, behavioral response, and regional transportation management strategies were evaluated to examine the relationship between these various variables and the performance of the transportation system.Thus, speed time-space plots and volume time-space plots on critical corridors as well as the Macroscopic measure of performance (Macro MOE) can be draw for different scenarios. A further analysis demonstrated that evacuation demand as well as management strategies, such as contraflow and staged evacua-tion play important role in the evacuation process. A mathematical method, curve fitting was applied to extract network performance function in respond to demand and management strategies from the simulation results. Such function, not only can be used to estimate net-work performance given demand and management plans, but also be able to analyze how network performance would change in case one or two parameters changed. Such infor-mation can be used for emergency authorities and government stake holders to make deci-sions in the object of help evacuees get out of endangered areas safely as fast as possible.It is expected that the knowledge and results gained from this research will be adaptable to evaluate other locations with different road network, population, transportation, and haz-ard threats.更多还原