基于小波分解和SVD降噪的齿轮箱特征提取研究  

Gearbox feature extraction research based on wavelet decomposition and SVD noise reduction

作　　者：朱云博[1] 冯广斌[1] 孙华刚[1] 

机构地区：[1]军械工程学院军械技术研究所,石家庄050003

出　　处：《起重运输机械》2011年第4期37-40,共4页Hoisting and Conveying Machinery

摘　　要：齿轮箱故障诊断的关键是对故障特征的提取。利用小波变换的多分辨特性,将齿轮箱振动信号进行分解及单支重构,获取原信号在不同频段上分布的详细信息,找出对应系统特征频率的尺度,并应用奇异值分解的方法对该尺度下的重构信号进行进一步的降噪处理,从中成功提取出信号的特征分量。The fault feature extraction is the key in the gearbox fault diagnosis.The paper decomposes vibration signal of the gearbox and reconstructs the single branch by using the multi-resolution property of the wavelet transformation,obtains the detailed distribution information of original signal in different frequency bands,finds out the scale corresponding to the system characteristic frequency,and applies further noise reduction processing to the reconstructed signal under the scale with SVD method,thus to successfully extract the characteristic component.

关 键 词：齿轮箱 故障诊断 特征频率 小波变换 奇异值分解 

分 类 号：TH132.41[机械工程—机械制造及自动化]