中国机械工程

• 智能制造 • 上一篇    下一篇

基于全矢融合与多维经验模态分解的滚动轴承退化过程频谱结构研究

马艳丽;金兵;张学欣;韩捷   

  1. 郑州大学振动工程研究所,郑州,450001
  • 出版日期:2017-07-25 发布日期:2017-07-26

Study on Frequency Spectrum Structure for Rolling Bearing Degradation Based on Full Vector Fusion and MEMD

MA Yanli;JIN Bing;ZHANG Xuexin;HAN Jie   

  1. Research Institute of Vibration Engineering,Zhengzhou University,Zhengzhou,450001
  • Online:2017-07-25 Published:2017-07-26

摘要: 为了识别滚动轴承退化过程,提出一种多维经验模态分解和全矢融合相结合的方法。首先对不同状态的多通道信号同时进行多维经验模态分解,得到一系列多元固有模态函数分量,然后利用互相关系数准则选取最敏感的一阶固有模态函数分量进行全矢包络分析来提取信号的特征。为了验证该方法的有效性,分别对模拟信号和实际信号进行了分析。结果表明此方法在出现故障时,能够很好地表征频谱结构的变化;随着故障严重程度的增加,频谱结构变得复杂,且呈现出了规律性。

关键词: 多维经验模态分解, 全矢融合, 频谱结构, 轴承退化

Abstract: In order to recognize degradation processes of rolling bearings, a method combined MEMD with full vector fusion was proposed. Firstly, multi-channel synchronous analyses of vibration signals in different conditions were dealt with MEMD to obtain a series of multivariate intrinsic mode function(IMF) component. Then the most sensitive IMF component was selected by cross-correlation coefficient criterion to do full vector envelope analysis for extracting signal features. Analysis with simulated signal and actual signal was done to test effectiveness of the method respectively. It indicates that the method may reveal changes of frequency structure when faults appear, and shows the regularity that the more serious of the fault conditions is, the more complex of frequency spectrum structure becomes.

Key words: multivariate empirical mode decomposition(MEMD), full vector fusion, frequency spectrum structure, bearing degradation

中图分类号: