Ouput-only vibration analysis of buildings in operational conditions

目前，由于传感技术和分析程序的进步，最受欢迎的实验 大型结构的模态试验是环境振动试验。环境振动分析导致必要的 建筑物的结构特性称为振型特性：振型、固有频率和阻尼 比率。这些信息对于跟踪结构完整性的未来变化是必不可少的，作为其 建立基准并帮助监测结构的健康状况。在正常运行期间进行环境振动测试 对结构而言，因此有不同的振源：风、微震、人类活动。 这些振动源的存在需要非常仔细地分析记录的弱信号以 能够准确地推导出模态特性。现有的频率和时间域方法有 在分析噪声信号、密集模式和阻尼比方面存在缺陷。他们是 工程师使用的复杂性，以及它们的优缺点也没有明确说明。这个 该项目的目标是开发一种新的混合频域/时间域方法来改进模型 身份识别过程。本项目的目标概述如下： I)对现有方法进行详细的文献审查，以清楚地确定其利弊。 二)开发能够处理噪声较高的环境信号的新算法，允许使用较少的 灵敏的低成本传感器。 Iii)所开发的算法应该能够更有效地处理密集模式。 IV)开发一种现场验证时间记录的简单方法，以便技术人员能够重新进行 测试是否检测到问题。 这项技术转让对SENFEFAKE公司和世界各地的工程界至关重要。

Currently, owing to advances in sensing techniques and analysing procedures, the most popular experimental modal test for large structures is ambient vibration testing. Ambient vibration analysis leads to the essential structural characteristic of buildings called modal properties: mode shapes, natural frequencies and damping ratios. These information are essential to keep track of future changes in the integrity of a structure, serve as its baseline and help monitoring health of structures. Ambient vibration test is performed during normal operation of the structure, therefore there are different source of vibrations: wind, micro-tremors, human activities. Existence of these numbers of vibration sources requires very careful analysis of the recorded weak signals to be able to accurately derive modal properties. Current existing frequency and time domain methods have shortcomings with regard to analysis of noisy signals, closely space modes and damping ratios. They are complex to be used by engineers, and also their advantages and disadvantages are not clearly specified. The goal of the proposed project is to develop a new mix frequency / time domain method to improve the modal identification process. The goals of this project are summarized as follow: i) Detailed literature review of exiting methods to clearly identify their pros and cons. ii) To develop new algorithm that can deal with ambient signals with higher noise, allowing to use less sensitive low-cost sensors. iii) The developed algorithm should be able to handle the closely-spaced modes more efficiently. iv) To develop a simple method for on-site verification of time records so that the technician can re-conduct the test if a problem is detected. This technology transfer is essential for Sensequake Company and engineering community around the world.