基于小波包分解和模糊聚类的网格结构损伤诊断方法

Damage diagnosis of spacial grid structures based on wavelet packet decomposing and fuzzy clustering

本文针对网格结构的特点,利用其在快速正弦扫频激励下的动力响应,基于小波包分解频带能量分析和模糊聚类法,提出了一种适合于网格结构的损伤诊断方法。利用两个传感器损伤信息的联合诊断,有效地解决了使用一个传感器无法检测出对称损伤的难题。利用有限元模拟,应用上述方法对一单层球面网壳结构进行了损伤诊断。结果表明,对网壳杆件较小程度的损伤,子频带能量成分变化敏感,用其构造诊断向量和建立损伤样本库,并进行模糊聚类和目标识别,基本可以正确地识别出网壳杆件的损伤位置和程度。同时该方法测试时间短,使用设备少,有可靠的理论基础,具有一定的工程应用参考价值。

In this paper, according to the characteristics of spacial grid structures, a method of damage presented by the dynamic response under fast sine sweep motivation based on the analysis of frequency of wavelet packet decomposition and fuzzy cluster method （FCM）. The problem of detecting symmetric diagnosis is band energy damage with only one sensor is effectually resolved by combined the diagnosis with two sensors. The method is tested by the fi- nite element analysis for a single-layer latticed shell. The results show that the energy components of sub-frequency bands are sensitive to the small-degree damage. Energy components are used as a diagnosis vector to establish damage sample base, fuzzy cluster and target identification, hence the damage location and degree for different elements can be identified basically. This approach has some advantages such as short detecting time, less equipments, reliable theoretical basis and engineering value.