基于功率谱密度矩心的轴承性能退化特征研究

滚动轴承性能退化指标作为研究其性能退化过程中极其重要的一环,良好的退化特征指标能够直观地表征轴承的退化情况并实现对轴承的退化预警。鉴于此,提出了将功率谱密度的矩心值作为性能退化特征,通过集合经验模态分解(EEMD)滤波降噪,在傅里叶变换的基础上,对原有的功率谱密度函数的重心进行计算,并采用高斯模型拟合优化该特征的表征能力。采用美国辛辛那提大学的轴承全寿命实验数据作为分析对象,以验证所提方法的可靠性。实验结果显示,该特征可以良好地反映轴承的性能退化状态。

Research on Bearing Performance Degradation Based on Centroid of Power Spectral Density

The index of rolling bearing performance degradation is an extremely important part in the study of its performance degradation process.A good degradation characteristic feature can intuitively characterize bearing degradation and realize early warning of bearing degradation.The centroid value of power spectral density(PSD)is used as a performance degradation feature,and the noise is filtered by ensemble empirical model decomposition(EEMD).Based on the Fourier transformation,the center of gravity of the original power spectral density is calculated.Then the researcher uses Gaussian models to fit and optimize the characterization ability of this feature.Finally,the whole bearing life data of the University of Cincinnati is used as the analysis object to verify the reliability of the proposed method.Experimental results show that this feature can reflect the performance degradation state of the bearing well.