噪声与振动控制

›› 2012, Vol. 32 ›› Issue (2): 108-111.DOI: 103969/j.issn.1006-1355-2012.02.026

• 6.信号处理与故障诊断 • 上一篇    下一篇

一种EMD端点延拓算法及其在齿轮故障诊断中的应用

李月仙1,韩振南1,高建新1,杨昊林2   

  1. （ 1. 太原理工大学 机械工程学院， 太原 030024；2. 忻州供电分公司， 山西 忻州 035100 ）
  • 收稿日期:2011-07-04 修回日期:2011-08-05 出版日期:2012-04-18 发布日期:2012-04-18
  • 通讯作者: 李月仙

An Endpoint Extension Algorithm of EMD and Its Application in Gear Fault Diagnosis

LI Yue-xian1,HAN Zhen-nan1, GAO Jian-xin1,YANG Hao-lin2   

  1. ( 1. College of Mechanical Engineering, Taiyuan University of Technology, Taiyuan 030024, China; 2. Power Supply Filiale in Xinzhou, Xinzhou 035100, Shanxi China )
  • Received:2011-07-04 Revised:2011-08-05 Online:2012-04-18 Published:2012-04-18
  • Contact: LI Yue-xian

摘要： 经验模态分解（Empirical Mode Decomposition，简称EMD）是目前分析非平稳信号的有效方法，但在分解过程中由于存在着严重的端点效应而影响分解结果。利用灰色理论的GM(1，1)模型来预测和修正端点的延拓值，使得在筛分过程中可以得到有效的三次样条插值包络线，从而抑制端点效应。通过对齿轮故障信号的分解验证，结果表明该方法是合理的。

关键词: 振动与波, 经验模态分解, GM(1, 1)模型, 边缘效应, 齿轮故障诊断

Abstract: EMD is an effective method for non-stable signal analysis, but the endpoint effect exists in the decomposition process. In this article, the GM (1, 1) model was used to predict and modify the extension values at the endpoints, which made the cubic interpolation envelope more effective in sifting process. In this way, the endpoint effect was effectively suppressed. The result of gear fault signal analysis shows that this method is correct and effective.

Key words: vibration and wave, EMD, GM (1, 1)model, edge effect, gear fault diagnosis

中图分类号: