Identification of Anomaly in Structures Based on Locally Controlled Dynamic Inputs

摘要快速准确地检测在役结构的异常是工程中的一大挑战。传感器和智能材料技术的最新进展为克服当前耗时费力的检测方法提供了有希望的机会。基于传感器的技术成功的关键很大程度上取决于传感器测量值如何与异常的大小和位置的物理量相关联。尽管具有挑战性，但如果生成数据的输入得到很好的控制，基于传感器的系统的数学复杂性将大大降低。这导致了一个基本的数学问题：给定由受控输入产生的有限传感器数据，确定对象的局部条件。因此，研究人员开展了一项研究，以开发一个数学框架，用于使用由局部控制的动态激励产生的分布式传感器测量，以及根据测量结果识别异常位置和大小的适当算法来检测结构中的异常。调查期间将进行分析和实验工作。本研究的主要任务包括：诊断信号的产生、信号的询问与解释、信号的实现与验证。虽然将使用简单的联片试验来验证结果，但数学框架是基本的，并将允许工程师探索新的数学公式，以解释任何复杂系统的数据。例如，该框架可用于监测飞机结构的疲劳裂缝，探测地下管道的腐蚀裂缝，为航天器的早期故障提供预警，或在大地震后询问桥梁或建筑物的完整性。此外，询问算法将为传感和监测技术在广泛的工程领域的应用提供有用的工具。这个项目是由CMS在数学-工程学接口计划下支持的。

AbstractCMS-0424141Chang, Fu-KuoStanford UniversityRapid and accurate detection of anomaly in structures while in service is major a challenge in engineering. Recent advances in sensor and smart materials technologies provide promising opportunities to overcome current time-consuming and labor-intensive inspection methods. The key to the success of the sensor-based technologies depends strongly on how the sensor measurements can be correlated with the physical quantity in terms of size and location of the anomaly. Although challenging, the mathematical complexity of the sensor-based systems becomes significantly reduced if the inputs to generate the data are well controlled. This leads to a fundamental mathematical issue: Given limited sensor data resulting from controlled inputs, identify local condition of an object.Therefore, an investigation is undertaken to develop a mathematical framework for detecting anomaly in structures using distributed sensor measurements generated from locally controlled dynamic excitations as well as appropriate algorithms to identify the location and size of the anomaly based on the measurements. Both analytical and experimental work will be conducted during the investigation. The major tasks to be performed for the study include: Diagnostic Signal Generation, Signal Interrogation and Interpretation, and Implementation and Verification. Although simple coupon tests will be used to verify the results, the mathematical framework is fundamental and shall allow engineers to explore new mathematical formulations for data interpretation of any complex systems. For instance, the framework could be applied to monitor fatigue cracks in aircraft structures, to detect corrosion cracks for underground pipelines, to provide early warning of incipient failure in spacecraft, or to interrogate the integrity of bridges or buildings after major quakes. Furthermore, the interrogation algorithms will provide useful tools for readily applying sensing and monitoring techniques for a broad range of engineering fields.This project is supported by CMS under the Math-Eng Interfacing Initiative.