GOALI: Hybrid Dynamic Feedback for Partially Autonomous Cooperative Active Safety Systems

该奖项是根据2009年《美国复苏和再投资法案》(公法111-5)提供资金的。该奖项的研究目标是为交通路口的部分自动化协作主动安全系统开发算法和实验演示。智能交通系统(ITS)的这一实例利用全球定位系统和无线通信技术来设计车载预警和控制系统。这些系统应该保证交通安全，同时适应人类司机的存在和不完善的感官信息。这项研究将产生动态协调控制算法，这些算法可以随着车辆数量的增加而扩展，并可以考虑到人类的行为。该研究方法依赖于状态信息不完全的混合动态模型的基于模型的控制，以及利用交通系统的自然单调动态特性。这种动力学的保序结构使区间方法的使用能够降低协调控制任务的复杂性。可交付成果包括可扩展算法、在安娜堡丰田全尺寸实验设施上实现此类算法的软件系统的开发、在同一设施上的实验验证、工程专业学生教育以及本科生的研究经验。如果成功，本研究结果将为未来的协作主动安全系统提供高效的算法。这些系统有可能极大地减少事故和死亡人数，并增加交通流量。这项研究将通过在安娜堡丰田技术中心实验设施上的实施，进一步提供向技术的具体过渡。结果将通过会议，特别是通过向公众进行试验性演示来传播。从事这个项目的研究生和本科生将从学术界和产业界之间的合作中受益，他们将接触到具体的问题，并了解他们对适合形式分析的形式的抽象过程。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). The research objective of this award is to develop algorithms and experimental demonstrations for partially automated cooperative active safety systems at traffic intersections. This instance of intelligent transportation systems (ITS) exploits global positioning systems and wireless communication technology to design on-board warning and control systems. These systems should guarantee traffic safety while adapting to the presence of human drivers and imperfect sensory information. The research will result in dynamic coordinated control algorithms that are scalable with the number of vehicles and can account for human behavior. The research approach relies on model based control of hybrid dynamical models subject to imperfect state information and on exploiting the naturally monotone dynamics of traffic systems. This order preserving structure of the dynamics enables the employment of interval methods to reduce the complexity of the coordinated control task. Deliverables include scalable algorithms, the development of a software system implementing such algorithms on the Ann Arbor TOYOTA full scale experimental facility, experimental validation on the same facility, engineering student education, and research experience for undergraduate students.If successful, the results of this research will provide efficient algorithms for the cooperative active safety systems of the future. These systems have the potential of dramatically decreasing the number of accidents, fatalities, and of increasing traffic flow. This research will further provide a concrete transition to technology through the implementation on the Ann Arbor TOYOTA Technical Center experimental facility. The results will be disseminated through conferences and in particular through experimental demonstrations to the public. Graduate and undergraduate students working on this project will benefit from the collaboration between academia and industry by being exposed to concrete problems and to the process of their abstraction to forms suitable for formal analysis.