运用分量信号小波能量法识别时变结构损伤

Damage Detection of Time-varying Strutures Using Wavelet Energy of Component Signals

在采用解析模态分解定理提取分量信号的基础上,对一阶分量信号进行复Morlet小波变换,并定义损伤前后一阶小波能量比值指标来识别时变结构的损伤位置,然后从损伤位置处的响应信号出发,引入滑动时间窗思想,提出归一化一阶小波能量变化指标,来预测结构损伤的演化过程。采用一个刚度突变和线性变化的三层剪切型建筑结构数值算例对提出的损伤指标进行验证,结果表明:该指标能够有效识别结构的损伤位置和时变损伤。

Complex Morlet wavelet transform is first conducted on the first-order mode component signal which is extracted based on analytical modal decomposition theorem. Then the ratio of the first-order mode wavelet energy indexes with the damage and without the damage is defined to detect the damage location of time-varying structures. Based on the decomposed component signal at the structural damage locations, the normalized first-order mode wavelet energy change is proposed to predict the structural damage evolution by introducing the concept of moving time window. A numerical example of three-story shear building with abrupt stiffness reduction and linear stiffness reduction is simulated to verify the effectiveness of the presented method. The results demonstrate that the defined index can effectively detect the damage locations and the time-varying damage of the structures.