CAREER: Multi-Resolution Model and Context Aware Information Networking for Cooperative Vehicle Efficiency and Safety Systems

美国每年约有3万人死于道路交通事故，220万人受伤。这一问题与交通拥堵造成的巨大经济损失更为严重。协同车辆效率与安全（CVES）系统的进步有望显著降低交通运输的人力成本和经济成本。然而，这种系统的大规模部署受到重大技术和科学差距的阻碍，特别是在实现协作车辆的实时和高精度态势感知方面。这个职业发展项目旨在通过发展自动化系统协调控制的基本信息网络方法来缩小这些差距。这些方法将基于建模知识传播的创新概念。此外，这个项目的教育部分整合了跨学科的信息物理系统（CPS）主题，将自动化网络系统的设计纳入研究生和本科生的培训模块。为了实现鲁棒性运行，CVES系统要求每辆车对其他协调车辆的状态具有可靠的实时感知。该项目通过开发一种模型和上下文感知的多分辨率信息网络方法，解决了对强大的面向控制的态势感知的关键需求。该方法是发展模型通信及其衍生的多分辨率网络的新概念。上下文感知模型通信依赖于模型的传输和同步（例如，随机混合系统结构和参数），而不是原始测量。这使得CVES动态模型在网络上的高保真同步成为可能。多分辨率网络概念是通过模型的可伸缩表示实现的。多分辨率模型允许网络内模型保真度适应可用的网络资源。结果表明，CVES对网络服务的可变性具有鲁棒性。CVES的成功部署，即使只是部分部署，也将通过减少交通事故和提高效率，带来显著的社会效益。该项目将通过解决其可扩展性挑战，实现大规模CVES部署。此外，该项目开发的方法对新兴的自动驾驶汽车至关重要，它们也有望在通信网络上协调行动。动态模型网络同步知识传播的基础研究成果将广泛应用于智能电网等其他CPS领域。该项目的教育部分将针对CPS研究人员和工程师进行智能交通和能源系统方面的培训。

Every year around 30,000 fatalities and 2.2 million injuries happen on US roads. The problem is compounded with huge economic losses due to traffic congestions. Advances in Cooperative Vehicle Efficiency and Safety (CVES) systems promise to significantly reduce the human and economic cost of transportation. However, large scale deployment of such systems is impeded by significant technical and scientific gaps, especially when it comes to achieving real-time and high accuracy situational awareness for cooperating vehicles. This CAREER project aims at closing these gaps through developing fundamental information networking methodologies for coordinated control of automated systems. These methodologies will be based on the innovative concept of modeled knowledge propagation. In addition, the educational component of this project integrates interdisciplinary Cyber-Physical Systems (CPS) subjects on the design of automated networked systems into graduate and undergraduate training modules. For robust operation, CVES systems require each vehicle to have reliable real-time awareness of the state of other coordinated vehicles. This project addresses the critical need for robust control-oriented situational awareness by developing a multi-resolution information networking methodology that is model- and context-aware. The approach is to develop the novel concepts of model communication and its derived multi-resolution networking. Context-aware model-communication relies on transmission and synchronization of models (e.g., stochastic hybrid system structures and parameters) instead of raw measurements. This allows for high fidelity synchronization of dynamical models of CVES over networks. Multi-resolution networking concept is enabled through scalable representations of models. Multi resolution models allow in-network adaptation of model fidelity to available network resources. The result is robustness of CVES to network service variability. The successful deployment of CVES, even partially, will provide significant societal benefits through reduced traffic accidents and improved efficiency. This project will enable large scale CVES deployment by addressing its scalability challenge. In addition, methodologies developed in this project will be crucial to emerging autonomous vehicles, which are also expected to coordinate their actions over communication networks. The fundamental research outcomes on knowledge propagation through network synchronization of dynamical models will be broadly applicable in other CPS domains such as smart grid. The educational component of this project will target training of CPS researchers and engineers on subjects in intelligent transportation and energy systems.