EEMD降噪方法在飞机强度试验异响识别中的应用  

Application of EEMD Noise Reduction Method to Abnormal Sound Recognition of Aircraft Strength Test

作　　者：韩凯[1] 冯建民[1] 贺谦[1] 孙志强[2] 

机构地区：[1]中国飞机强度研究所,全尺寸飞机结构静力/疲劳航空科技重点试验室,陕西西安710065 [2]国防科技大学机电工程与自动化学院,湖南长沙410073

出　　处：《工程与试验》2016年第3期9-13,74,共6页Engineering and Test

基　　金：航空基金(20150981006,20140937001)资助

摘　　要：针对飞机强度试验中异常信号被背景噪声淹没、提取信号特征困难这一问题,提出了结合相关系数的集合经验模态分解方法(EEMD)降噪方法。首先对信号进行EEMD降噪,然后根据相关系数筛选出用于重构信号的IMF分量,提取特征值,最后运用支持向量机进行分类辨识。通过与几种降噪法进行比较,结果表明,结合相关系数的EEMD降噪方法优于其它降噪方法,更适用于充满噪声的全尺寸飞机强度试验中。Aiming at the problems that the abnormal sound signal is buried in the background noises in aircraft strength test and it is difficult to obtain available abnormal sound signal feature,the noise reduction method based on ensemble empirical mode decomposition（ EEMD） and correlation coefficient is proposed. Firstly,abnormal sound signal is decomposed by EEMD and a set of intrinsic mode function components are obtained. Filtering the components according to the correlation coefficient,the selected IMF components are used to reconstruct the signal. Lastly,energy feature extracted from a number of IMFs could serve as input vectors of support vector machine.Practical examples show that the method has obvious advantages compared with other noise reduction methods and can be used in aircraft strength test with the strong background noise.

关 键 词：集合经验模态分解方法 降噪 相关系数 特征提取 支持向量机 

分 类 号：V216.54[航空宇航科学与技术—航空宇航推进理论与工程]