自适应MED结合EMD诊断滚动轴承早期故障.PDF

第35 卷 第6 期 噪 声 与 振 动 控 制 Vol 35 No.6 2015 年12月 NOISE AND VIBRATION CONTROL Dec. 2015 文章编号：1006-1355(2015)06-0159-04 自适应MED 结合EMD 诊断滚动轴承早期故障 刘尚坤，唐贵基 （华北电力大学 能源动力与机械工程学院，河北 保定，071003 ） 摘 要：针对最小熵解卷积（MED ）降噪效果受滤波器长度影响问题，提出一种利用步长迭代算法和包络谱熵检验 准则的自适应MED 降噪方法，并结合经验模态分解（EMD ）提取滚动轴承微弱故障特征。首先利用自适应MED 降噪 方法对原信号进行最优降噪处理，然后通过EMD 将降噪信号分解为若干个本征模函数（IMF）分量，再选取峭度值最大 的IMF 进行包络谱分析，根据包络谱中故障特征频率实现故障诊断。仿真信号和实测信号分析结果表明其优于基于 EMD 的包络解调方法。 关键词：振动与波；最小熵解卷积；经验模态分解；包络谱熵；滚动轴承；故障诊断 中图分类号：TH 133.3 文献标识码：A DOI 编码：10.3969/j.issn. 1006-1335.2015.06.034 Incipient Fault Diagnosis of Rolling Bearings Based on Adaptive MED and EMD LIU Shang-kun , TANG Gui-ji ( School of Energy, Power and Mechanical Engineering, North China Electric Power University, Baoding 071003, Hebei China ) Abstract : Since the de-noised result of the minimum entropy deconvolution (MED) was affected by the filter length, an adaptive MED de-noising method using step iterative algorithm and envelope spectrum entropy criterion was proposed. This method was combined with the empirical mode decomposition (EMD) to extract the fault frequency of rolling bearing ’ s weak faults. First of all, the fault signal of the rolling bearing was de-noised by the adaptive MED method. Then, the de- noised signal was decomposed into several intrinsic modal function (IMF) components by EMD. Finally, the bearing ’s fault characteristic frequency was extracted by the envelope spectrum analysis of the selected IMF with the maximum kurtosis value. The analysis results of the simulation sign