AMC-SS Stochastic Modeling and Simulation for Traffic Flow

提案ID：0606807 PI：Alexandros SopasakisInstitution：马萨诸塞州阿默斯特大学题目：交通流的AMC-SS随机建模与仿真摘要复杂的动态系统，如交通流，其特征是在不同的时间尺度上共存的相位。此外，随机扰动，虽然很小，可以对整个系统的行为有显着的影响。虽然许多物理现象是由大量变量的相干但先验未知的运动引起的，但目前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。我们提出了一个层次的交通流模型，可以提供一个更好的方式来预测交通流观测。具体来说，我们采用了一种新的，能源驱动，随机噪声模型，以描述不同的复杂阶段的车辆交通流。该模型将能够再现交通流的混沌和难以解决的高浓度regimes.Modeling车辆交通流的行为是一个具有挑战性的，当前和开放的问题，需要立即解决。建议的交通流模型可以是一个突破，几个跨学科的交通相关的研究领域，使更准确和潜在的更快的模拟现实的交通流包含大量的相互作用的车辆。交通的连贯描述可以导致更好的预测策略，在我们的能源，环境和安全问题上有直接和间接的好处。

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.