Cost-effective Rehabilitation of Civil Infrastructure using Limited Sensors

即使是设计得最好、建造得最仔细的结构，比如建筑物和桥梁，最终也会变弱。无论结构是由于环境条件或使用而随着时间的推移而失效，还是由于地震或强风等自然灾害而突然失效，失效都会对人类和社会造成重大影响。在结构使用过程中检测损伤，既可以避免安全风险，又可以最大限度地降低暂停进入评估损伤的经济成本。最近，基于振动的损伤检测（VBDD）方法作为结构健康监测的一个分支，在民用基础设施的修复工作中显示出了希望。在VBDD中，通过检测振动测量中的异常症状来评估结构中存在的损坏及其严重程度，然后使用适当的决策工具进行修复。然而，由于传感器的高成本及其安装问题，在大型结构的每个关键位置都安装传感器并不总是可行的。拟议的研究将开发智能和经济高效的技术，使用有限数量的振动传感器进行早期损伤诊断和修复。

Even the best-designed, most carefully built structures, such as buildings and bridges, eventually weaken. Whether a structure fails over time due to environmental conditions or use, or suddenly, because of a natural disaster such as an earthquake or strong wind, failure has significant human and societal consequences. Detecting damage while a structure is in use both helps avoid safety risks and minimizes the economic costs of suspending access to assess damage. Recently, vibration based damage detection (VBDD) methods, an offshoot of structural health monitoring, have shown promise in efforts to rehabilitate civil infrastructure. In VBDD, the presence of damage and its severity in a structure is assessed through detecting unusual symptoms in vibration measurements, and then rehabilitation is undertaken using appropriate decision making tools. However, because of the high cost of sensors and their installation issues, it is not always feasible to install sensors at every critical location in a large-scale structure. The proposed research will develop smart and cost-effective technology to perform early damage diagnosis and rehabilitation using limited numbers of vibration sensors. Current tools for damage detection and maintenance use a wide range of signal processing methods. Recent development of blind source separation (BSS) by the applicant as a powerful class of signal decomposition methods makes them a promising candidate for VBDD. Recently, BSS has been successfully demonstrated in linear structures in a wide range of challenging situations. In the proposed research, BSS algorithms will be expanded to develop sophisticated damage detection algorithms for deteriorating (i.e., nonlinear) structures using a fewer number of vibration measurements. Various decision making approaches, relating to sensor densities, distributed sensing, optimal sensor placement, retro-fitting and control strategies, will be considered. These algorithms will result in cost-effective rehabilitation and technology-enabled maintenance of deteriorating civil structures. Such a multidisciplinary research and development will advance efforts to attain sustainable and resilient infrastructure. Researchers and students will develop diverse skills and experience in fields such as structural engineering, mechanical engineering, systems design and decision analysis. This program will develop several research projects with a tangible impact on the smart infrastructure community.