期刊文献+

基于孪生网络的货车滚动轴承异常检测

下载PDF
导出
摘要 轴承是铁路列车走行部中的关键组成部分,其工作环境复杂,容易发生故障,不能保障列车运行安全。根据经验对运营条件下的货车故障轴承进行识别,缺乏智能检测技术。本文基于孪生网络(Siamese)和局部离群因子(Local outliers factor, LOF)算法实现了复杂轴承特征的异常检测。在真实运营条件下验证了方法的有效性。结果显示,利用傅里叶变换的频域信号识别效果比时域信号更显著;与传统方法相比,孪生网络和分类算法结合的方法对故障和正常轴承特征的分类效果更好;LOF和一类支持向量机(One Class SVM,OCSVM)作为分类器均能实现4.4%和5.6%提升,而LOF算法分类效果比OCSVM更好。
作者 赵京平
出处 《中国新技术新产品》 2024年第12期28-30,共3页 New Technology & New Products of China
  • 相关文献

参考文献3

二级参考文献31

  • 1Paya B A, Esat I I .Artificaial Neural Network Based Fault Diagnostics of Rotating Machinery Using Wavelet Transforms sa a Preprocessor .Mechanical Systems and Signal Processing 1997 .11(5):751-765.
  • 2Huang N E, et al .The Empirical Mode Decomposition and the Hilbert Spectrum for Nonlimear and Non-Stationary Time Series Analysis .Proc .R .Soc.Lond. A,1998,454:903-995.
  • 3Lou X S, Loparo K A. Bearing fault diagnosis based onwavelet transform and fuzzy inference [ J ]. MechanicalSystems and Signal Processing, 2004, 18(5) : 1077 - 1095.
  • 4Jiang Q S, Jia M D, Hu J Z, et al. Machinery fault diagnosisusing supervised manifold learning[ J] . Mechanical Systemsand Signal Processing, 2009,23(7) ; 2301 -2311.
  • 5Sun W X, Chen J, Li J Q. Decision tree and PCA-based faultdiagnosis of rotating machinery [ J ]. Mechanical Systems andSignal Processing, 2007, 21(3) :1300-1317.
  • 6Yu J B. Bearing performance degradation assessment usinglocality preserving projections and Gaussian mixture models[J]. Mechanical Systems and Signal Processing, 2011, 25(7): 2573 -2588.
  • 7Janjarasjitt S, Ocak H, Loparo K A. Bearing conditiondiagnosis and prognosis using applied nonlinear dynamicalanalysis of machine vibration signal [ J ]. Journal of Sound andVibration, 2008,317(1 -2) : 112-126.
  • 8Brand M, Oliver N, Pentland A. Coupled hidden Markovmodels for complex action recognition [ C ] //IEEE ComputerSociety Conference on Computer Vision and PatternRecognition, San Juan, PR, USA, 1997, 994-999.
  • 9Bernhard Scholkopf, Alexander Smola, Mtiller K R.Nonlinear component analysis as a kernel Eigen value problem[J]. Neural Computation, 1998: 1299 - 1319.
  • 10朱义.基于CHMM的设备性能退化评估方法研究[D].上海:上海交通大学,2008.

共引文献236

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部