一种自适应提取有效信号的滚动轴承故障诊断方法  

A Fault Diagnosis Method for Rolling Bearings based on Adaptive Extraction of Effective Signals

作　　者：邢欣 崔亚辉[1] 刘晓琳[1] 王增杰 李龙龙 XING Xin;CUI Yahui;LIU Xiaolin;WANG Zengjie;LI Longlong(College of Mechanical and Precision Instrument Engineerin,Xi'an University of Technology,Xi’an 710048,China)

机构地区：[1]西安理工大学机械与精密仪器工程学院,西安710048

出　　处：《噪声与振动控制》2018年第2期150-153,161,共5页Noise and Vibration Control

摘　　要：针对目前故障诊断中从大量信号中难以自适应选取有效信号的问题,提出一种将小波包分解与shannon熵相结合自适应选取有效信号的方法。对故障信号进行小波包分解,计算频带熵值以量化信号的复杂程度。以熵值大小为指标,找寻小波包最大熵所在的一段频率信号,以此信号为有效信号进行小波包信号重构。将重构后信号进行EMD分解,对得到的IMF分量进行Hilbert包络谱分析,有效分离和突出故障频率。实验研究表明采用该方法自适应选取的有效信号能够保证所提取轴承故障特征频率的有效性和直观性,使故障诊断的实时性得到增强。Currently,it is difficult to adaptively extract the effective signals from a large number of signals.This paper proposes a method to select the effective signals by combining wavelet packet decomposition with Shannon entropy.First of all,the wavelet packet decomposition is carried out on the faulty signals to calculate the band entropy to quantify the complexity of the signals.With the value of the entropy as the index,the frequency signal corresponding to the maximum entropy of the wavelet packet is found,which is used as the effective signal for reconstruction of the wavelet packet signal.Then,the reconstructed signal is decomposed by EMD and the obtained IMF component is analyzed by Hilbert envelope spectrum.Finally,the fault frequency is effectively separated and highlighted.Through the experimental study,it is shown that the method can effectively select the effective signal,which can guarantee the validity and intuition of the characteristic frequency of the bearing fault,and improve the real-time performance of the fault diagnosis.

关 键 词：振动与波 故障诊断 自适应 小波包最大熵 EMD包络谱 

分 类 号：TH165[机械工程—机械制造及自动化]