基于改进的局部特征尺度分解的结构损伤识别  

Structural Damage Identification based on Improved Local Characteristic-scale Decomposition

作　　者：任宜春 刘鸿杨 Ren Yichun;Liu Hongyang

机构地区：[1]长沙理工大学土木工程学院,湖南长沙410114

出　　处：《重庆建筑》2020年第11期39-42,共4页Chongqing Architecture

摘　　要：针对Hilbert-Huang变换中经验模态分解和Hilbert谱分析的不足,应用改进的局部特征尺度分解和归一化正交法来识别工程结构中的损伤。该文通过一个三层线性剪切型框架结构数值模拟分析,对该方法的有效性进行了证明。最终得出结论:改进的局部特征尺度分解方法在处理原信号进行分解时,具有更好的正交性;通过归一正交法得到的瞬时频率图,能够清晰地判断出何时结构发生了损伤;对结构受损前后所计算得到的特征内秉尺度分量归一化能量比的差值进行比较,准确地找到结构损伤发生的位置。Due to the shortcomings of Empirical Mode Decomposition and Hilbert analysis in Hilbert-Huang transform,the improved local characteristic-scale decomposition and normalized quadrature were used to identify damages in engineering structures.The effectiveness of this method was confirmed through numerical simulation analysis of a three-story linear shear frame structure.The conclusion is that the Improved Local Characteristic Decomposition has better orthogonality when processing the original signal;the instantaneous frequency diagram obtained by normalized quadrature can clearly judge when the structure has been damaged;the difference of the normalized energy ratio of the intrinsic scale component within the feature before and after damage can be identified and the location of structural damage can be spotted as well.

关 键 词：自适应时频分析 改进的局部特征尺度分解 内禀尺度分量 损伤识别 

分 类 号：TU317.1[建筑科学—结构工程]