AMC-SS Stochastic Modeling and Simulation for Traffic Flow

交通流 AMC-SS 随机建模与仿真

• 批准号: 0836699
• 负责人: Alexandros Sopasakis
• 金额: $ 4.17万
• 依托单位: University of North Carolina at Charlotte
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-12-01 至 2010-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0836699&HistoricalAwards=false
• 关键词: AMC; SS; Stochastic; Modeling; Simulation

Proposal ID: 0606807PI: Alexandros SopasakisInstitution: University of Massachusetts AmherstTitle: AMC-SS Stochastic Modeling and Simulation for Traffic FlowAbstractComplex dynamical systems, such as traffic flow, are characterized by coexisting phases acting at different temporal scales. Further, random perturbations, although small, can have dramatic effects on the overall system behavior. While many physical phenomena result from the coherent but a priori unknown motion of a large number of variables most of the current models require that individual motion of the particles is resolved. Such approaches lead to difficulties in obtaining predictions that are dependable or timely. We propose a hierarchy of traffic flow models that can provide a better way of predicting traffic flow observables. Specifically we employ a novel, energy driven, stochastic noise model in order to describe the different complex phases of vehicular traffic flow. The model will be capable of reproducing the flow of traffic for the chaotic and difficult to resolve higher concentration regime.Modeling the behavior of vehicular traffic flow is a challenging, current and open problem demanding an immediate solution. The proposed traffic flow model can be a breakthrough for several cross-disciplinary traffic related research areas by enabling much more accurate and potentially faster simulation of realistic traffic streams containing a large number of interacting vehicles. The coherent description of traffic can lead to better prediction strategies with direct and indirect benefits in our energy resources, environmental and safety issues.

交通流的随机建模与仿真摘要交通流等复杂动力系统具有在不同时间尺度上相互作用的特征。此外，随机扰动虽然很小，但可以对整个系统的行为产生巨大的影响。虽然许多物理现象是由大量变量的相干但先验的未知运动引起的，但目前的大多数模型都要求解决粒子的单个运动。这种方法导致难以获得可靠或及时的预测。我们提出了一种交通流模型的层次结构，它可以提供一种更好的预测交通流观测值的方法。具体来说，我们采用了一种新的能量驱动的随机噪声模型来描述车辆交通流的不同复杂阶段。该模型将能够再现交通流的混乱和难以解决的高浓度状态。车辆交通流行为的建模是一个具有挑战性的、当前的和开放的问题，需要立即解决。所提出的交通流模型可以通过更准确和潜在更快地模拟包含大量相互作用车辆的现实交通流，从而成为多个跨学科交通相关研究领域的突破。交通的连贯描述可以带来更好的预测策略，在我们的能源、环境和安全问题上有直接和间接的好处。