基于局部均值分解和极值延拓的旋转机械故障特征提取方法  

Rotating Machinery Fault Diagnosis Method Based on the Local Mean Decomposition and Extreme Points Extension

作　　者：孟宗[1] 王亚超[1] 王晓燕[1] 

机构地区：[1]燕山大学河北省测试计量技术及仪器重点实验室,河北秦皇岛066004

出　　处：《计量学报》2014年第5期469-473,共5页Acta Metrologica Sinica

基　　金：国家自然科学基金（51105323）

摘　　要：局部均值分解对非平稳、非线性故障信号进行平稳化处理时表现出特有的分析能力，能够有效获得故障信号的时频特征，然而局部均值分解过程中存在的端点效应严重影响信号的分解效果。针对这一问题，提出了一种基于局部均值分解和极值延拓的旋转机械故障提取方法。首先采用极值延拓方法处理信号的两个端点，左、右端点均分别延拓2个极大值和2个极小值，然后对延拓后的信号进行局部均值分解，提取信号中包含的故障特征。仿真结果表明，经过极值点延拓后，局部均值分解过程中的端点效应得到了有效抑制，最后以轴承内圈故障为例在实验平台进行了实验研究，实验结果表明，该方法能有效提取出旋转机械故障特征。Local mean decomposition shows unique analysis ability when smoothing process the nonstationary, nonlinear fault signal. It can effectively obtain time-frequency characteristics of fault signal. However, the end effect existing in the process of the local mean decomposition seriously affects the signal decomposition result. In order to solve this problem, a method of rotating machinery fault extraction based on the local mean decomposition and extreme points extension is proposed. Firstly, using the method of extreme points extension to process the two endpoints of the signal. The left and right endpoints are respectively extended two maximums and two minimums. Then, using the method of local mean decomposition to decompose the signal with extension, and extract fault features it contains. The simulation results show that, after extreme points extension, the end effect in the process of local mean decomposition have been effectively suppressed. Finally, with the bearing inner race fault as an example experimental study was carried out in experimental platform. The experimental results show that this method can effectively extract the fault characteristics of rotating machinery.

关 键 词：计量学 局部均值分解 端点效应 旋转机械 

分 类 号：TB936[一般工业技术—计量学]