CRII: Algorithms and Methodologies for Real-Time Decision-Making of Mission-Critical Structures Experiencing High-Rate Dynamics

这个项目的重点是研究方法，以便能够对经历高速率动态的关键任务结构系统进行实时决策。经历高速动力学的关键任务结构的例子包括高超声速飞行器、航天器、弹道组件和主动爆炸缓解。通过增强这些结构的生存能力，提供及时的指导修正，并根据不断变化的条件采用任务目标/结果，使这些结构能够进行实时决策，从而提高了任务成功率。此外，该项目开发的方法提高了在极端动态环境中运行的结构的健壮性、安全性和商业可行性。为经历高速率动力学的结构开发算法和方法符合国家利益，并完成了NSF的使命：促进科学进步；促进国家健康、繁荣和福利；或确保国防安全。这个项目提供了研究经验，并指导了一群不同和包容的学生，并引入了一个研究生水平的课程，内容是复杂系统的代理建模。在高速动态环境中运行的结构系统可能会经历结构的突然和未建模的塑性变形，这可能进一步导致电子设备、传感器和/或脆弱的有效载荷损坏。这项研究的重点是使实时决策模块能够采取纠正行动。为了实现这一目标，开发了一个并行化框架，该框架使结构系统能够被分解成其组成组件，其中每个组件都可以被监视、建模并推断到未来。一旦估计了每个组件的退化轨迹，它们就被重新组合成一个单一的系统级模型，用于实时决策。该项目被组织为两个研究推动力和一个实验设计挑战。推力1研究代理建模技术，目标是开发能够在所需时间限制内收敛的组件级模型。这些代理模型使用从密集传感器网络获得的数据来生成数据驱动的损伤敏感特征，这些特征可以用作组件级预测的退化参数。《推力2》探索并制定了实时组件级别预测的方法。然后，这些组件级别的预测被重新组合到单个结构模型中，该模型用于制定潜在的纠正措施。最后，实验设计挑战使用高速率动态测试基准来验证开发的算法和方法。该项目由高级网络基础设施办公室(OAC)和既定的激励竞争研究计划(EPSCoR)联合资助。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project focuses on investigations into methodologies to enable real-time decision-making for mission-critical structural systems experiencing high-rate dynamics. Examples of mission-critical structures that experience high-rate dynamics include hypersonic vehicles, space crafts, ballistics packages, and active blast mitigation. Enabling real-time decision-making for these structures increases mission success rates by enhancing the structure's survivability, providing on-time guidance corrections, and adopting the mission goals/outcome to changing conditions. Additionally, the methodologies developed by this project increase the robustness, safety, and commercial viability of structures operating in extreme dynamic environments. The development of algorithms and methodologies for structures experiencing high-rate dynamics serves the national interest and fulfills the NSF's mission: to promote the progress of science; to advance the national health, prosperity and welfare; or to secure the national defense. This project provides research experience and mentors a diverse and inclusive group of students, and introduces a graduate level class on surrogate modeling of complex systems.A structural system operating in a high-rate dynamic environment can experience sudden and unmodeled plastic deformation of the structure that may further lead to damaged electronics, sensors, and/or delicate payloads. This research focuses on enabling a real-time decision-making module to take corrective actions. To achieve this goal, a parallelization framework is developed that enables a structural system to be decomposed into its constituent components where each component can be monitored, modeled, and extrapolated into the future. Once the degradation trajectories for each component have been estimated, they are recombined into a single system-level model to be used for real-time decision making. The project is organized into two research thrusts and one experimental design challenge. Thrust 1 investigates surrogate modeling techniques with the goal of developing component-level models that can converge within the required time constraints. These surrogate models use data obtained from dense sensor networks to generate data-driven damage-sensitive features that can be used as the degradation parameters for component-level prognostics. Thrust 2 explores and formulates methodologies for real-time component-level prognostics. These component-level predictions are then recombined into a single structural model that is used to develop potential corrective actions. Lastly, the experimental design challenge validates the developed algorithms and methodologies using a high-rate dynamic test bench.This project is jointly funded by Office of Advanced Cyberinfrastructure (OAC) and the Established Program to Stimulate Competitive Research (EPSCoR).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.