CAREER: Development of Nonlinear Modeling Tools for Analysis, Simulation, and Structural Health Monitoring

我们的许多民用基础设施正处于严重退化的状态。该项目将开发基于有关关键结构系统振动响应特征的有限信息来评估其健康状况的方法，为工程师提供一套可靠的工具来评估其持续生存能力，并在有限的预算内确定维修和维护的优先顺序。与传统的结构健康监测方法(数据以离线/批处理模式分析)不同，该项目将通过强调算法效率标准，将涉及实时损伤检测的自适应系统识别的新方法开发为一套面向应用的工具。这些工具将通过分阶段的实验计划进行评估，包括数值模拟、实验室实验和全面评估。在PI先前工作的基础上，这项研究将展示非线性识别技术在结构健康监测方面以及在结构工程/工程力学的更广泛领域中的应用潜力。研究的目标是：1)开发混合方法，绕过用于损伤检测目的的非线性结构系统的非参数和参数建模的困难；2)开发用于结构健康监测和主动控制的非线性实时识别技术，并寻求从替代的传感器技术中受益；3)评估从低水平激励或仅通过响应测量(即未知激励)识别结构动力学的局限性；4)评估从稀疏的振动传感器阵列推断损伤的局限性，并开发使工程师能够容易地检测结构损伤的经济有效的方法；5)将这些工具应用于其他工程挑战，如MEMS、土壤动力学和航空航天。为完成研究项目，开展了一个综合教育项目。该教育项目有三个主要方面：1)开发一个实验室设施，用于研究和学生参与的可重复非线性动态测试--通过使用主动控制来合成非线性行为来实现重复性；2)开发一门研究生级别的健康监测课程；以及3)暑期推广计划，通过教学、实地考察、实验室演示和动手测试，让纽约市中心的贫困高中生参与工程研究环境-所有这些都是与哥伦比亚大学的双重发现中心合作的。

Much of our civil infrastructure is in a severe state of degradation. This project will develop methods to assess the health of critical structural systems based on limited information about their vibration response signature, providing engineers with a set of reliable tools to evaluate their continued viability and prioritize repair and maintenance within limited budgets. Unlike conventional structural health monitoring methods (where data is analyzed in an off-line/batch mode) the project, through an emphasis on algorithm efficiency criteria, will develop new approaches in adaptive system identification involving real-time damage detection into a set of application-oriented tools. These tools will be evaluated through a phased experimental program, involving numerical simulation, laboratory experimentation, and full-scale evaluation.Building on the PI's prior work, the research will show the potential of nonlinear identification techniques for application in the structural health monitoring context, and in broader fields of structural engineering/engineering mechanics. The research objectives are: 1) to develop hybrid methodologies, circumventing difficulties associated with nonparametric vs. parametric modeling of nonlinear structural systems for damage detection purposes; 2) to develop nonlinear real-time identification techniques for structural health monitoring and active control, and seek to benefit from alternative sensor technologies; 3) to evaluate limitations in identification of structural dynamics from low-level excitation or from response measurements only (i.e., unknown excitation); 4) to evaluate limitations of inferring damage from vibration sensor arrays that are sparse, and develop cost-effective methodologies that enable engineers to readily detect structural damage; and 5) to apply these tools to other engineering challenges, e.g., MEMS, soil-dynamics and aerospace.Complementing the research project, an integrated educational project is undertaken. This educational project has three main aspects: 1) the development of a lab facility for repeatable nonlinear dynamic testing for research and student participation - repeatability achieved through use of active control to synthesize nonlinear behavior; 2) the development of a graduate-level course on health monitoring; and 3) a summer outreach program to engage underprivileged, inner-city, New York high-school students in the engineering research environment through teaching, field-trips, laboratory demonstration, and hands-on testing - all in collaboration with Columbia University's Double Discovery Center.