基于内积延拓LMD的机床滚动轴承故障诊断方法研究

A machine tool rolling bearing fault diagnosis method based on integral waveform extension LMD

针对机床加工中常见的滚动轴承故障特性,提出了一种基于内积延拓LMD的故障诊断方法。该方法针对LMD方法本身的端点效应缺陷,采用内积延拓的方法对分析信号左右两端点进行匹配延拓以抑制端点效应。首先建立原始信号的特征波形和相似波形并计算其对应的积分值;其次通过计算相似波形和特征波形积分差值并从信号的左端开始迭代运算以寻找最优匹配波形;最后将最优匹配波形延拓到对应的信号左右两侧以完成极值点的端点延拓。该方法充分考虑了信号内部波形的特点以及内部趋势运行规律,使得延拓后的波形可以很好地保持信号左右两端原有的趋势。实验采用滚动轴承故障信号进行分析,实验结果表明该方法可以很好地抑制LMD端点效应,并可以有效地提取滚动轴承故障特征实现故障诊断,可以运用到机床轴承的故障诊断中。

Aimed at the fault feature of rolling bearing during the machining process by machine tool,this paper proposes a new fault diagnosis method based on integral waveform extension Local Mean Decomposition( LMD). This method introduces the integral waveform extension method to extent the left and right side of the analyzed signal,to overcome the end effect of pure LMD. Firstly,the feature waveform and similar waveform are established and the corresponding integral values are calculated. Secondly,the difference values between these integral values are calculated to obtain the minimum one,which is important in finding the optimal matching waveform. At last,the left and right optimal matching waveforms are extended on the left side and right side of the analyzed signal,respectively. This method considers the inner characteristics of the analyzed signal,and could keep the signal trend better. The rolling bearing experiment results proves the effectiveness of the new method on both suppressing LMD end effect and diagnosing the bearing faults,and the new method could be used for machine tool fault diagnosis.