首页|CEEMDAN和盲源分离在轴承复合故障诊断中的应用

CEEMDAN和盲源分离在轴承复合故障诊断中的应用

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
滚动轴承的复合故障信号中往往含有多个特征信息及背景噪声,为更高效实现故障信息的提取,提出一种基于具有自适应白噪声的完备集成经验模态分解(CEEMDAN)和盲源分离的滚动轴承复合故障特征提取方法.对实验所获取的故障数据进行CEEMDAN分解,得出一组固有模态函数(IMF),利用加权峭度因子选取其中有效IMF重构信号,再将重构的信号进行BSS分离.对分离出的信号做解调包络分析,从其解调谱中提取故障信号的特征频率.结果证明了此方法可以有效地分离轴承的内外圈故障,使故障特征更易被提取.
Application of CEEMDAN and Blind Source Separation in Bearing Compound Fault Diagnosis
The composite fault signal of rolling bearing often contains multiple feature information and background noise.In or-der to extract fault information more efficiently,a feature extraction method of rolling bearing composite fault based on comple-mentary ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and blind source separation is proposed.A set of intrinsic mode function(IMF)is obtained by CEEMDAN decomposition of the fault data obtained in the experiment.The effec-tive IMF reconstruction signal is selected by using the correlation kurtosis factor,and then the reconstructed signal is separated by blind source separation.The demodulation envelopment analysis of the separated signals is done,and the characteristic frequen-cies of the fault signal are extracted from the demodulation spectrum.The results show that this method can effectively separate the inner and outer ring faults of bearing and make the fault features easier to be extracted.

Rolling BearingsComplementary Ensemble Empirical Mode Decomposition with Adaptive NoiseBlind Source SeparationCorrelation Kurtosis FactorFeature Extraction

古莹奎、林忠海、刘平

展开 >

江西理工大学机电工程学院,江西 赣州 341000

滚动轴承 自适应白噪声的完备集成经验模态分解 盲源分离 加权峭度因子 特征提取

国家自然科学基金资助江西省自然科学基金

6196301820181BAB202020

2024

机械设计与制造
辽宁省机械研究院

机械设计与制造

CSTPCD北大核心
影响因子：0.511
ISSN：1001-3997
年,卷(期)：2024.(3)
  • 15