ERI: Towards Safe Aviation Autonomy: A Risk-bounded Planning Framework for Dynamical Systems under Uncertainties

该奖项全部或部分由2021年美国救援计划法案（公法117-2）资助。该工程研究启动（ERI）奖支持研究，以实现在日益自主的环境中的安全操作，包括国家空域系统（NAS）。无人机交付系统、城市空中机动性和商业太空运输的发展和采用，使得将这些模式整合到当前的NAS中变得至关重要。当拥挤的空域被大量的无人机交通和太空发射期间的预留飞机危险区域共享时，不隔离和有效的操作是非常具有挑战性的。提高NAS的自主性可以通过增加有限空域中的流量密度来帮助缓解拥塞，但需要高度的安全操作保证。该项目将研究这些多种空中交通模式的集成的风险有界规划和运营，并将提供一个基本的理论，一套模型，并在动态和不确定的环境，包括NAS的风险有界规划的解决方案算法的自治系统社区。该项目将为不确定环境下的动力系统风险有界规划奠定方法基础。在此框架内，该项目将研究风险约束规划算法和分布式计算方法，以集成NAS内的自主操作。该项目将提供三种创新技术来支持NAS环境中的规划和操作：（I）数据驱动的方法，用于动态数据的不确定性的有效表征;（II）一套基于机会约束和不确定性动态系统凸近似的风险有界规划算法;（III）具有多尺度方法的分布式计算框架，其可以有效地解决计及一般不确定性的机会约束模型。这些方法将用于开发风险约束规划算法和动态概率地理围栏，以支持NAS中的自主空中交通整合。该项目将支持研究生，并加强在圣地亚哥州立大学，西班牙裔服务高等教育机构的研究生课程。PI还将使用一个模拟平台来吸引K-12和本科生，特别是那些来自代表性不足的群体的学生，在航空航天工程，信息科学和人工智能领域。该奖项反映了NSF的法定使命，并通过使用基金会的智力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 award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2).This Engineering Research Initiation (ERI) award supports research to enable safe operations in the increasingly autonomous environment that comprises the national airspace system (NAS). Development and adoption of drone delivery systems, urban air mobility, and commercial space transportation make it critical to integrate these modes into the current NAS. Unsegregated and efficient operations are very challenging when congested airspace is shared by high volume drone traffic and reserve aircraft hazard areas during space launches. Increasing autonomy in the NAS can help relieve congestion by increasing traffic density in limited airspace but demands high assurance of safe operations. This project will study risk-bounded planning and operations for the integration of these multiple air traffic modes and will provide the autonomous systems community with an underlying theory, set of models, and solution algorithms for risk-bounded planning in the dynamical and uncertain environment that comprises the NAS. The project will develop the methodological foundations for risk-bounded planning of dynamical systems under uncertain environments. Within this framework, the project will study risk-bounded planning algorithms and a distributed computing approach to integrate autonomous operations within the NAS. The project will deliver three innovative techniques to support planning and operations in the NAS environment: (I) a data-driven method for efficient characterization of uncertainty with dynamical data; (II) a suite of risk-bounded planning algorithms based on chance constraints and convex approximation for dynamical systems under uncertainties; (III) a distributed computing framework with a multi-scale approach that can efficiently solve the chance-constrained models accounting for generic uncertainties. These methods will be employed to develop risk-bounded planning algorithms and a dynamic, probabilistic geofence to support autonomous air traffic integration in the NAS. The project will support graduate students and enhance the graduate program at San Diego State University, a Hispanic Serving Higher Education Institution. The PI will also use a simulation platform to engage K-12 and undergraduate students, especially those from underrepresented groups, in the areas of aerospace engineering, information science, and artificial intelligence.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.