基于冗余2代小波局部梯度谱熵的轴承故障诊断  

Rolling bearing fault diagnosis based on redundant second generation wavelet and local gradient spectrum entropy

作　　者：张园[1] 李力[2] 

机构地区：[1]三峡大学科技学院,湖北宜昌443002 [2]三峡大学机械与材料学院,湖北宜昌443002

出　　处：《制造技术与机床》2015年第6期105-108,共4页Manufacturing Technology & Machine Tool

摘　　要：用改进的冗余第2代小波包将滚动轴承振动信号分解为多个频带载波,以互相关系数最大原则确定最优载波,求其局部梯度及局部梯度谱图,并提出局部梯度谱熵指标对试验工况滚动轴承进行故障分类。结果表明:最匹配的故障轴承载波信号局部梯度功率谱能找到其对应特征频率及其倍频。局部梯度谱熵指标大小顺序基本为:外圈剥落最大,其次是内圈剥落、滚动体剥落和正常,能较好区分不同试验转速的轴承信号。By improved Redundant Second generation wavelet,rolling bearing vibration signals are decomposed into several carriers,and the best carrier is assured by highest cross correlation coefficient principle,work out its local gradient and local gradient spectrum,then local gradient spectrum entropy is proposed to classify the experiment rolling bearings of working condition. The result shows that the fault feature frequencies and its frequencies＇ doubling can be found on power spectrum of the best carrier local gradient. The order of local gradient spectrum entropy is the outer race pitting is the biggest,followed by inner ring pitting,rolling body exfoliation,and normal,it can preferably distinguish the signals of different experiment speeds.

关 键 词：滚动轴承 改进冗余第2代小波 局部梯度 局部梯度谱熵 

分 类 号：TH133.3[机械工程—机械制造及自动化]