中国机械工程

• 机械基础工程 • 上一篇    下一篇

基于改进奇异值分解滤波和谱峭度的滚动轴承故障诊断

孟宗;刘子涵;吕蒙   

  1. 燕山大学河北省测试计量技术及仪器重点实验室,秦皇岛,066004
  • 出版日期:2020-10-25 发布日期:2020-10-29
  • 基金资助:
    国家自然科学基金资助项目(52075470,61873227);
    河北省自然科学基金资助项目(E2019203448);
    中央引导地方科技发展基金资助项目(206Z4301G)

Fault Diagnosis for Rolling Bearings Based on Improved Singular Value Decomposition and Spectral Kurtosis

MENG Zong;LIU Zihan;LYU Meng   

  1. Key Laboratory of Measurement Technology and Instrumentation of Hebei Province,Yanshan University,Qinhuangdao,Hebei,066004
  • Online:2020-10-25 Published:2020-10-29

摘要: 针对含噪信号的有效奇异值个数难以确定的问题,提出了一种改进的奇异值分解降噪方法——奇异值累积法。该方法通过计算奇异值的实际下降值与奇异值平均下降速度累积量的差值,并取该差值最大值点的位置作为有效奇异值的分界点来确定有效奇异值的个数。在此基础上,提出了一种基于奇异值累积法与快速谱峭度的滚动轴承故障诊断方法。采用奇异值累积法对原信号进行降噪处理,然后利用快速谱峭度确定滤波器中心频率及带宽,通过分析频段包络谱中明显的频率成分来诊断故障。该方法可以有效去除信号中的噪声,使得到的峭度值所反映的故障冲击更接近实际情况。对含内圈、外圈故障的滚动轴承实验数据进行分析,实验结果表明,相比快速谱峭度的故障诊断方法,该方法具有更好的故障识别效果。

关键词: 滚动轴承, 故障诊断, 奇异值分解, 有效奇异值, 快速谱峭度

Abstract: For the problems that is difficult to determine the number of effective singular values of noisy signals, an improved SVD method—singular value accumulation method was proposed. The difference between the actual falling and the cumulative value of the average falling of the singular value was calculated, and the position of the maximum point of the difference was taken as the boundary point of the number of effective singular values to determine the effective singular value number. A fault diagnosis method for rolling bearings was proposed based on singular value accumulation method and fast spectral kurtosis. Firstly, the singular value accumulation algorithm was used to denoise the original signals and the center frequency and bandwidth of the filter was determined by the fast spectral kurtosis. At last, fault features might be diagnosed by analyzing obvious frequency components in envelope spectrum. The proposed method may effectively denoise the signals, so that the fault impacts reflected by the obtained kurtosis value are closer to the actual situation. By analyzing the experimental data of rolling bearings with inner and outer ring faults, the experimental results show that the proposed method has better fault identification effectiveness than that of the fault diagnosis method with the fast spectral kurtosis.

Key words: rolling bearing, fault diagnosis, singular value decomposition(SVD), effective singular value, fast spectral kurtosis

中图分类号: