考虑温度及操作变量影响的结构损伤预警研究

发布时间：2018-05-25 01:18

本文选题：AR模型 + 因子分析　；参考：《青岛理工大学》2013年硕士论文

【摘要】：在基于振动的损伤识别中，结构损伤会引起损伤特征参数的改变，温度和操作变量的影响也会使这些特征参数发生变化，而且这种变化往往会掩盖结构自身损伤所引起的特征参数变化，从而无法区分结构损伤指标的变化是由损伤引起的还是由温度和操作变量引起的，因此，有必要进行温度及操作变量影响下的结构损伤识别研究。本文中以时间序列中的AR模型参数为损伤特征参数，利用多元统计分析中的因子分析和主成分分析去除温度及操作变量（质量荷载）对损伤特征参数的影响，结合标准差控制图对结构进行损伤预警，主要内容如下：（1）综述了去除温度因素及操作变量影响的方法，指出其应用的范围和存在的问题。（2）简述了AR模型及因子分析和主成分分析的基本理论。（3）针对温度因素及操作变量对结构损伤特征参数的影响，本文提出了利用因子分析及主成分分析两种方法去除温度及操作变量对损伤特征参数（AR模型参数）的影响，并结合标准差控制图进行结构损伤预警。（4）本文中温度影响分为两种：结构整体温度均匀变化和结构温度不均匀变化，两种温度影响下的损伤工况各不相同，数值模拟结果表明：结构温度不均匀变化对损伤识别的影响高于结构整体温度均匀变化产生的影响，而因子分析及主成分分析均可以有效的去除温度的影响，并且去除温度影响后的损伤特征参数通过标准差控制图可以准确进行损伤预警。（5）结构质量变化的影响通过改变海洋平台顶板的对角点的堆积质量来实现，在质量变化的同时对结构引入损伤，数值模拟结果表明：因子分析及主成分分析均可以有效的去除操作变量（质量变化）的影响，，并且去除质量变化影响后的损伤特征参数通过标准差控制图可以准确进行损伤预警。
[Abstract]:In Vibration-based damage identification, structural damage will cause changes in damage characteristic parameters, temperature and operating variables will also cause changes in these characteristic parameters. Moreover, this kind of change will often cover up the characteristic parameter change caused by the damage of the structure itself, so it is impossible to distinguish whether the change of the damage index of the structure is caused by the damage or by the temperature and the operating variable. It is necessary to study the structural damage identification under the influence of temperature and operating variables. In this paper, the AR model parameters in time series are taken as damage characteristic parameters, and the effects of temperature and operation variables (mass load) on damage characteristic parameters are removed by factor analysis and principal component analysis in multivariate statistical analysis. Combined with the control chart of standard deviation, the damage warning of the structure is carried out. The main contents are as follows: The methods of removing temperature factors and operating variables are reviewed, and the application scope and existing problems are pointed out. The AR model and the basic theories of factor analysis and principal component analysis are briefly introduced. 3) aiming at the influence of temperature factors and operating variables on structural damage characteristic parameters, this paper proposes two methods to remove the effects of temperature and operation variables on the damage characteristic parameters and AR model parameters by factor analysis and principal component analysis. Combined with the control chart of standard deviation, structural damage warning is carried out. In this paper, the influence of temperature is divided into two kinds: uniform change of structure temperature and non-uniform change of structure temperature, and the damage conditions are different under the influence of two kinds of temperature. The results of numerical simulation show that the influence of non-uniform change of structure temperature on damage identification is higher than that of the uniform change of temperature of the whole structure, and factor analysis and principal component analysis can effectively remove the influence of temperature. And the damage characteristic parameters after removing temperature influence can be accurately forewarned by standard deviation control chart. 5) the influence of structural mass change is realized by changing the stacking mass of diagonal points on the roof of offshore platform, which causes damage to the structure at the same time of mass change. The results of numerical simulation show that both factor analysis and principal component analysis can effectively remove the influence of operation variables (mass change). And the damage characteristic parameters after removing the influence of mass change can be accurately forewarned by standard deviation control chart.
【学位授予单位】：青岛理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TU317

【参考文献】