Collaborative Research: A Dynamic Disruption Prediction System for Transportation Networks at a Road-Segment Level of Granularity

该资助将发展理论基础，并在经验上测试一种新的框架的可行性，该框架用于在道路段粒度级别的极端事件中实时监测和道路运输网络中断预测。最近的自然灾害表明，道路交通网络的失败主要是由于不寻常的交通模式造成的意外堵塞。这些中断对疏散、救援和恢复行动等应急管理过程产生不利影响。对于一个成功的应急管理系统来说，及时预测中断是至关重要的，但尽管有这一关键需求，应急管理过程仍然缺乏交通系统的实时网络监测和预测方法。道路网络的实时监测和中断预测为开发主动应急管理系统创造了机会，从而实现高效、快速和成功的救援和恢复行动。更有效和成功的应急管理行动的好处将传递给公众，并将提高通勤者和其他基础设施用户的生活质量、健康和福祉。最终，该项目将有助于发展可持续和有弹性的城市和社区，即使在极端事件的压力下也能正常运作。与这个项目有关的是，设想将教育和外联工作纳入本科和研究生课程。这项资助将为K-12，本科生和研究生，特别是少数民族和代表性不足的群体提供基于项目的学习机会。该研究项目旨在开展基础研究，以创建一种方法和框架，根据极端事件中检测到的异常交通模式，动态预测道路交通网络的中断。本项目设计的方法提供了一种具有路段粒度级别的整体网络方法，该方法在考虑整个网络的同时，也能够监控和预测每个路段的交通流量。具体的研究目标是：1)发现和建立时间交通流相互依赖关系的模型；2)利用瞬时交通数据在路段粒度级别实时有效地监控交通流，以检测异常交通模式；3)通过整合网络相互依赖关系和识别出的异常交通模式来预测交通中断。该项目的成果将为建立主动应急管理系统奠定基础，从而提高救援和恢复行动的效率和成功率。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 grant will develop theoretical foundations and empirically test the feasibility of a novel framework for real-time monitoring and disruption prediction of road transportation networks during extreme events at a road-segment level of granularity. Recent natural disasters have shown that road transportation networks mainly fail due to unexpected gridlocks resulting from unusual traffic patterns. These disruptions adversely affect emergency management processes such as evacuations, rescue, and recovery operations. Prompt prediction of disruptions is vital for a successful emergency management system, but despite this critical need, emergency management processes still lack real-time network monitoring and prediction methods for transportation systems. Real-time monitoring and disruption prediction of road networks creates opportunities to develop proactive emergency management systems that lead to efficient, fast, and successful rescue and recovery operations. The benefits of more efficient and successful emergency management operations will be passed down to the public and will result in enhanced quality of life, health, and well-being for commuters and other infrastructure users. Ultimately, this project will contribute to developing sustainable and resilient cities and communities that can function properly even under the stress of extreme events. In connection with this project, educational and outreach efforts are envisioned for integration into undergraduate and graduate courses. This grant will provide an opportunity for project-based learning for K-12, undergraduate, and graduate students, especially minorities and under-represented groups. This research project aims to conduct fundamental research to create a methodology and framework that dynamically predicts disruptions in road transportation networks based on unusual traffic patterns detected during extreme events. The method designed in this project offers a holistic network approach with a road segment level of granularity that, while it considers the entire network, is also able to monitor and predict traffic flow in each road segment. The specific research objectives are to 1) discover and model temporal traffic flow interdependencies, 2) efficiently monitor traffic flows on a real-time basis at a road segment level of granularity using instantaneous traffic data to detect unusual traffic patterns, and 3) predict disruptions by integrating network interdependencies and the identified unusual traffic patterns. The outcomes of this project will set the stage for development of proactive emergency management systems that will result in more efficient and successful rescue and recovery operations.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.