论文部分内容阅读
针对结构健康监测中如何基于在线监测数据实现损伤诊断的问题,提出了一种利用时间序列分析ARMA模型的特征提取和损伤预警方法.首先对所有监测数据样本建立ARMA模型,以模型中AR部分参数的主成分矩阵构建Mahalanobis距离判别函数,提出了一种新的结构损伤敏感指标DDSF.然后,采用t-检验考察该指标在损伤前后是否存在显著性变化,从而可以有效地实现结构损伤预警.三跨连续梁数值算例表明,提出的结构损伤特征指标对结构的微小损伤具有敏感性,具备结构在线实时损伤预警的应用价值.
Aiming at the problem of structural damage monitoring based on on-line monitoring data in structural health monitoring, a method of feature extraction and damage pre-warning based on time series analysis is proposed.At first, ARMA model is established for all monitoring data samples, and the AR parameters A new structural damage sensitive index DDSF is proposed by using the principal component matrix of Mahalanobis distance discriminant function.Then the t-test is used to examine whether there is a significant change in the index before and after the damage, so that structural damage early warning can be effectively implemented. Numerical examples of the cross-continuous beams show that the proposed structural damage characteristics are sensitive to the small damage of the structure and have the application value of real-time structural damage warning.