小波变换在结构损伤识别特征提取中的应用

APPLICATION OF WAVELET TRANSFORMATION IN FEATURE EXTRACTION FOR STRUCTURE DAMAGE IDENTIFICATION

小波变换用来进行特征提取是近年来模式识别领域中较新的一个研究方向。不同于传统的特征提取方法,它借助信号处理领域常用的小波变换这一数学工具,对特征库进行再次加工,经过去粗取精而得到更利于分类的新的信息。针对结构损伤识别特征提取的特点,发展了小波变换对基本特征库进行分频段处理的双谱相辅分析技术,并开发了相应的软件系统,通过算例表明,该方法可以达到降低噪声,凸显结构损伤信息的目的。

Application of wavelet transformation in feature extraction is a newer subject recently in the field of modal identification.Based on wavelet transformation in the field of signal processing,the feature aggregate is reprocessed in order to get the new information for easy classification.According to the characteristics of feature extraction of structure da-mage identification,the double-spectrum analysis technology is put forward and the corresponding software system is deve-loped.Then an example is used to prove the effectiveness of this method and it shows that the proposed technology can reduce noise and display structure damage information.