CyberSEES: Enabling Sustainable Civil Infrastructure Using Interactive Formal Analytics for Structural Health Diagnosis

该项目寻求一种全面的方法，通过结构健康监测(SHM)和计算正式方法之间迄今尚未探索的交叉，同时考虑过程中的人工交互，来对关键基础设施进行实时资产管理。一个多学科团队将计算机科学家和在现实世界公路桥梁结构健康监测方面经验丰富的土木工程师聚集在一起。这项研究不同于传统的基础设施健康监测方法，通过应用可满足性模理论(SMT)求解器对基于力学的约束进行结构模型的直接求解。该项目的主要目标是验证这种正式的结构健康诊断方法的有效性。这项活动包括开发一个框架，以构建通用的和特定于结构的约束，并使用SMT对其进行逻辑编码，以形式地约束基于测量的结构模型的解空间，以满足对基于振动的监测至关重要的逆特征问题。还将对基于SMT的诊断结果的解释进行调查，初步研究交互式视觉分析的作用，利用有形的交互作用来增强从业者感知、理解和促进结构健康监测系统诊断的能力。维护桥梁和其他老化的民用基础设施的安全是一个具有挑战性的社会问题。结构健康监测提供了一条可持续的途径，使使用寿命最大化，并为我们昂贵的关键基础设施提供基于优化时间表的维护，而不会危及公共安全或商业和战略公路路线的功能。如果成功，该项目将为全面的诊断和预测框架奠定坚实的基础。这种智能基础设施模型可以直接翻译到其他广泛的网络物理系统中。为了促进数据共享和传播，实验室桥梁模型将作为开源资源分发给更广泛的科学和工程界，并通过NSF NEEShub项目仓库发布和维护实验细节、分析模型和数据集。将项目研究与教育相结合的活动包括开发课程模块、项目和案例研究。

This project seeks a holistic approach to real-time asset management of critical infrastructure through an as-yet unexplored intersection amongst structural health monitoring (SHM), and computational formal methods, while considering human interaction in the process. A multidisciplinary team brings computer scientists together with civil engineers experienced in structural health monitoring of real-world highway bridges. The research departs from the conventional approach to infrastructure health monitoring by direct solution of a structural model through application of a Satisfiability Modulo Theories (SMT) solver to mechanics-based constraints. The main objective of the project is to validate the efficacy of this formal approach to structural health diagnostics. This activity involves development of a framework to construct generalized and structure-specific constraints and encode them logically using SMT to formally bound the solution space of a measurement-based structural model satisfying the inverse Eigen problem that is central to vibration-based monitoring. Interpretation of the diagnostics result based on SMT will also be investigated, in a preliminary study of the role of interactive visual analytics utilizing tangible interaction to enhance the ability of practitioners to perceive, understand, and contribute to the diagnosis of the structural health monitoring system. Maintaining the safety of bridges and other aging civil infrastructure is a challenging societal problem. Structural health monitoring offers a sustainable pathway by maximizing the service life and informing optimized schedule-based maintenance of our costly critical infrastructure without jeopardizing public safety or the functionality of commercial and strategic highway routes. If successful, this project will create solid foundations for a comprehensive diagnosis and prognostication framework. This model for smart infrastructure is directly translatable to a broad spectrum of other cyber-physical systems. To facilitate data sharing and dissemination, the laboratory bridge model is to be distributed as an open-source resource to the broader scientific and engineering communities with experimental details, analytical models, and datasets published and maintained through the NSF NEEShub Project Warehouse. Activities to integrate the project research with education include development of course modules, projects, and case studies.