基于改进EMD和滑动峰态算法的滚棒轴承声发射信号故障特征提取

Fault characteristic extraction of AE signals based on improved EMD and sliding kurtosis algorithm ofrolling bearing

采用小波包对滚棒轴承声发射信号降噪,对降噪后的信号进行经验模式分解,选取特定本征模分量,采用滑动峰态算法提取其中的冲击分量,即提取滚棒轴承声发射信号的故障特征分量。改进的EMD方法剔除了某些虚假本征模分量,更准确地表征原始信号。通过仿真信号验证,成功提取了混合信号中的冲击分量,证明了该方法对冲击信号提取的有效性。对外圈故障的滚棒轴承声发射信号进行分析,滚棒轴承的故障特征频率及其倍频明显,对轴承故障的诊断具有重要的意义并可推广到航空发动机主轴轴承的故障诊断。

Using wavelet packetto denoiseacoustic emission signals,and the empirical mode decomposition （ EMD）method to decompose the denoising signal,to select the specific intrinsic modefunctions（ IMFs）,to extract the impact of components based on sliding Kurtosis algorithms,that is extracting the fault characteristics of the acoustic signals by rolling bearings. The improved EMD method removes the certain false specific intrinsic modefunctions,and presents more accurate characteristics of the original signals. Through simulation,the impact of mixed-signal componentsis successfully extracted. And the method of exacting impact signal sisproved to be effective. Through the analysis of the acoustic signals of the outer ring of rolling bearings,the fault characteristic frequencyand frequency multiplication of rolling bearing sareobviously displayed. The method can beextended to theaero-engine main bearings fault diagnosis.