Teager能量算子增强倒阶次谱提取轴承微弱故障特征
2016-01-06杨青乐,梅检民,肖静等
第一作者杨青乐男,硕士,1990年7月生
通信作者肖云魁男,博士,教授,博士生导师,1956年9月生
Teager能量算子增强倒阶次谱提取轴承微弱故障特征
杨青乐1,梅检民1,肖静2,张玲玲1,肖云魁1
(1.军事交通学院军用车辆系,天津300161; 2.军事交通学院图书馆,天津 300161)
摘要:针对变速器加速过程下轴承故障特征易于暴露难以提取问题,提出一种Teager能量算子增强倒阶次谱方法。计算加速过程等角度重采样信号的Teager能量算子,对Teager能量算子输出进行倒谱分析,获得Teager能量算子增强倒阶次谱。对加速过程滚动轴承外圈、内圈剥落故障信号进行分析,结果表明,Teager能量算子能有效增强冲击成分,抑制非冲击成分;倒阶次谱能从干扰中准确识别被增强的故障冲击特征,提取轴承微弱故障特征。
关键词:Teager能量算子;倒阶次谱;滚动轴承;故障诊断
收稿日期:2014-01-28修改稿收到日期:2014-04-13
中图分类号:TH165+.3文献标志码:A
基金项目:国家自然科学基金(U1434203, 51377136,51407147);中国铁路总公司科技研究开发计划(2013J010-B)
Weak fault feature extraction for bearings based on an order cepstrum enhanced with Teager energy operator
YANGQing-le1,MEIJian-min1,XIAOJing2,ZHANGLing-ling1,XIAOYun-kui1(1. Department of Military Automobile,Military Transportation University,Tianjin 300161, China;2.The Library of Military Transportation University,Tianjin 300161, China)
Abstract:Bearing’s fault features are easy to be exposed and difficult to be extracted during acceleration of a gearbox. In order to solve this problem, an order cepstrum enhanced with Teager energy operator was presented. Firstly, Teager energy operator of a signal during acceleration resampled with even angle-interval was calculated. Then, the output of Teager energy operator was analyzed with the order cepstrum method, the order cepstrum enhanced with Teager energy operator was obtained. The signals of exfoliation faults of outside and inside rings of a rolling bearing during acceleration were analyzed. The results showed that Teager energy operator can be used to effectively enhance impulse components and restrain non-impulse components; the order cepstrum can be used to identify fault impact features from interferences accurately and extract the weak fault features of the bearing.
Key words:Teager energy operator; order cepstrum; rolling bearing; fault diagnosis
变速器滚动轴承内、外圈发生局部(剥落、腐蚀等)故障时振动信号中会出现周期性冲击成分,轴承早期故障周期冲击振动时间短、时域能量微弱、频域带宽较宽,不易检测及提取。变速器变转速过程更能突出故障特征,但其它干扰及噪声背景更强,更难提取微弱特征。为有效提取轴承早期故障微弱特征,亟需探索有效的特征提取方法。
解调是轴承故障特征提取中应用较多的方法,通常需确定共振频带,且仅对信噪比较高的信号效果较好[1-2]。……
