The New Traffic Microscope- Measuring Microscopic Traffic Dynamics to Model and Control Freeway Traffic Congestion

这项NSF拨款将对高速公路交通动态产生更深入的了解，这将用于开发下一代交通模型，并最终有针对性地干预，以减少交通拥堵的负面影响。据估计，2017年交通拥堵造成的损失为1790亿美元。如果这项研究能够将这些成本降低几个百分点，它将产生高回报，对社会产生广泛影响。交通本身就很难研究，因为人们需要在大范围内进行精确的测量，这有点类似于能够在整个城市的卫星照片中阅读报纸。目前的交通动态和控制技术是基于低分辨率的数据，只能提供有关“平均车辆”的信息。美国国家科学基金会资助的研究最近的结果表明，推动更高的分辨率和了解单个车辆的相互作用对于推进理论和控制至关重要。本研究将(1)从本质上开发一个“显微镜”，通过大规模的经验数据集来观察这些单个车辆的相互作用；(2)使用新发现的动态来改进或取代当前的交通流模型；(3)使用这些交通流模型来开发有针对性的干预措施，以提高交通控制的有效性。高速公路交通动态的微观细节低于传统车辆检测器的分辨率。这项研究将使用新的测量技术来实证研究交通实际流动的微观本质，最终目标是建立更强大的交通流模型和更有效的交通控制。新的测量技术从常见的环路探测器中提取微观关系，从而以新的方式使用旧传感器的集合来达到在这个规模上无与伦比的分辨率水平。大量的高分辨率数据使研究能够分离出微观交通动态，而这些动态在此之前被噪声所掩盖。有了新的清晰度，微观关系将导致根本性的发现。研究将在三个层面上取得进展：(i)实证调查传统交通流量模型的可疑缺陷。（二）对于不足之处，探讨影响范围和基本机制的结构，以便更深入地了解依赖关系及其如何影响交通动态。（iii）最后，根据研究结果制定交通流模型，以准确捕捉动态，并探索如何利用这些见解改善交通控制。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This NSF grant will produce a deeper understanding of freeway traffic dynamics that will be used develop the next generation of traffic models and ultimately targeted interventions to reduce the negative impacts of traffic congestion. Traffic congestion was estimated to cost the US $179 Billion in 2017. If this research can reduce those costs by only a few percent it will have a high payoff with a broad impact to society. Traffic is inherently difficult to study because one needs fine measurements over a large scale, somewhat akin to being able to read a newspaper in a satellite photo of an entire city. The current state of the art in traffic dynamics and control is based on low resolution data that only provide information about the "average vehicle." Recent results from NSF sponsored research has shown that it is critical to push to higher resolution and understand the individual vehicle interactions to advance both theory and control. This research will (1) essentially develop a "microscope" to see these individual vehicle interactions across large scale empirical data sets, (2) use the newfound dynamics to improve or replace current traffic flow models, and (3) use these traffic flow models to develop targeted interventions that will improve the effectiveness of traffic control.The microscopic details of freeway traffic dynamics are below the resolution of conventional vehicle detectors. This research will use new measurement techniques to empirically study the microscopic nature of how traffic actually flows, with the ultimate goal of building more robust traffic flow models and more effective traffic controls. The new measurement techniques extract microscopic relationships from common loop detectors, thereby using a collection of old sensors in new ways to achieve a level of resolution is unrivaled at this scale. The massive amount of high resolution data allow the research to isolate microscopic traffic dynamics that heretofore were obscured by noise. With the new clarity the microscopic relationships will lead to fundamental discoveries. The research will progress on three levels: (i) Empirically investigate suspected deficiencies of conventional traffic flow models. (ii) For the deficiencies explore the bounds of influence and the structure of the underlying mechanisms to develop a deeper understanding of the dependencies and how they impact traffic dynamics. (iii) Finally, act on the findings by developing traffic flow models to accurately capture the dynamics and explore ways in which these insights can be used to improve traffic control.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.