%0 Journal Article
%T 基于改进信息图与MOMEDA的滚动轴承故障特征提取
%A 夏均忠
%A 于明奇
%A 白云川
%A 刘鲲鹏
%A 吕麒鹏
%J 振动与冲击
%D 2019
%X 为解决最大相关峭度解卷积存在的故障周期需要预先设置等问题，提出多点优化最小熵解卷积修正（Multipoint Optimal Minimum Entropy Deconvolution Adjusted, MOMEDA）用于增强轴承故障信号，并应用改进信息图降低噪声对其多点峭度谱的干扰。首先，通过引入轴承故障与正常状态下谱负熵的比值关系，优化信息图中平均谱负熵算法，提出基于滤波器组的改进信息图方法；其次，构建带通滤波器进行滤波降噪，并通过MOMEDA多点峭度谱识别故障周期；最后，应用MOMEDA增强滤波信号中的故障周期性脉冲成分，并通过平方包络谱提取微弱故障特征。试验表明，较之信息图等方法，改进信息图的降噪效果较为突出，可有效提高故障周期的识别度，实现MOMEDA自适应增强故障信号。</br>Abstract：To solve the selection problem of the faulted period in the maximum correlated kurtosis deconvolution method,multipoint optimal minimum entropy deconvolution adjusted (MOMEDA) was introduced to enhance the faulted signal.Furthermore,improved infogram was combined to denoise for MOMEDA multipoint kurtosis (MKurt) spectrum.First,the spectral negentropy ratio of both faulted and healthy signals was considered to improve the average spectral negentropy algorithm.As a result,improved infogram based on filter banks was proposed.Second,in order to correctly identify the faulted period in the MOMEDA MKurt spectrum,the proposed method was employed to construct an optimal band-pass filter for noise reduction.Third,the weak fault characteristics were extracted by means of the square envelope spectrum after the MOMEDA being applied to enhance periodic impulses in the filtered signal of rolling bearings.The results of measured fault signals show that the improved infogram is better on denoising.Moreover,it can effectively improve the faulted period's identification and help MOMEDA adaptively with enhancing faulted signals.
%K 滚动轴承
%K 特征提取
%K 信息图
%K 改进信息图
%K 多点优化最小熵解卷积修正&lt
%K /br&gt
%U http://jvs.sjtu.edu.cn/CN/abstract/abstract8266.shtml