摘要
针对齿轮箱轴承故障信号含有大量噪声而特征难以提取的问题。文章提出一种基于MCKD(最大相关峭度解卷积)和小波包熵值相结合的齿轮箱微弱故障信号提取方法。首先根据MCKD对故障信号进行降噪,突出信号中的有效冲击成分。然后进行小波包分解得到包含故障特征成分的末层节点信号,并以互相关系数-小波包熵值为准则对最后一层节点信号进行筛选并获取敏感节点信号,最后通过对敏感节点信号进行重构从而获得降噪后的轴承故障信号。实验结果表明该方法能够很好的滤除信号中的噪声并且准确地提取故障信号中的冲击成分,是对齿轮箱微弱故障特征提取的一种新方法。
For the problem that the fault signal of gearbox bearing contains a lot of noise,so that it is difficult to extract the features,this paper proposed a weak fault signal extraction way of gear box on the basis of combination of MCKD（maximum correlation kurtosis deconvolution） and wavelet packet entropy.Firstly,the fault signal was de-noised according to MCKD so as to highlight the effective component of the signal.Then the obtained terminal node of signal contained the characteristics of was decomposed by the wavelet packet,and the last node signals were filtered to obtain the sensitive node signals under the criterion of correlation coefficientwavelet packet entropy,finally,the fault signal of bearing after noise reduction was obtained through reconstruction the signal of the sensitive node.The verification of the simulated signal and the measured fault signal showed that the method can remove the noise well in the signal and extract the impact component of the fault signal accurately,which was a new method to extract the weak fault signal of gear box.
作者
杨思
朱爱华
姚德臣
杨建伟
白永亮
YANG Si;ZHU AI-hua;YAO De-chen;YANG Jian-wei;BAI Yong-liang(Beijing Key Laboratory of Service Performance of Urban Rail Transit Vehicles, Beijing University of Civil Engineering Architecture 100044, China;School of Mechanical and Electronic Control Engineering, Beijing Jiaotong University,Beijing 100044,China)
出处
《组合机床与自动化加工技术》
北大核心
2018年第5期131-134,138,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
中共北京市委组织部优秀人才(2012D005017000006)
国家自然科学基金资助项目(51605023)
长城学者培养计划项目(CIT&TCD20150312)
国家十三五重点研发计划城市轨道系统安全保障技术(2016YFB1200402)
关键词
MCKD
小波包熵
齿轮箱
滚动轴承
故障提取
MCKD
wavelet packet entropy
gearbox
rolling bearing
fault extraction
