谱峭度方法在柔性薄壁轴承故障特征频率提取中的应用

柔性薄壁轴承是一种特殊的轴承,工作时其内、外圈都受力变形为椭圆,这导致正常柔性薄壁轴承的振动信号存在周期性冲击,这种冲击背景使柔性薄壁轴承的故障特征提取更为困难。谱峭度方法是利用峭度对信号中的瞬态成分敏感这一特性,根据峭度值最大原则来准确提取共振带,克服了传统共振解调技术存在带通滤波器参数需要人工预先确定的缺陷。对自适应谱峭度,基于滤波器组的快速谱峭度以及基于短时傅里叶变换的快速谱峭度进行了对比分析,利用它们提取柔性薄壁轴承内、外圈故障特征频率,对这三种方法的特征提取效果进行了比较,结果表明自适应谱峭度方法更适合于提取柔性薄壁轴承的故障特征频率。

Application of spectral kurtosis method in fault feature frequency extraction of flexible thin-walled bearing

Flexible thin-walled bearing is a special type of bearing,and when working,its inner and outer rings are deformed to oval to cause there is periodic impact in its vibration signals.This impact background makes it more difficult to extract its fault features.The spectral kurtosis method uses a kurtosis characteristic of it being sensitive to transient components of signals.The resonance band is extracted accurately according to the principle of kurtosis being the maximum to overcome the defect of traditional resonance demodulation technology requiring manual pre-determination of band-pass filter parameters.Here,adaptive spectral kurtosis,fast kurtosis based on filter set and fast kurtosis based on short time Fourier transformation were analyzed contrastively,then they were used to extract fault feature frequency of inner and outer rings of a flexible thin-walled bearing,respectively.The feature extraction effects of these 3 methods were compared.The results showed that the adaptive spectral kurtosis method is more suitable for extracting fault feature frequency of flexible thin-walled bearings.