基于双谱分析的滚动轴承故障模式识别

Fault Pattern Recognition of Rolling Bearings Based on Bi-spectrum Analysis

当滚动轴承发生故障时,其产生的振动信号一般是包含较强噪声的非高斯和非线性信号。本文对高阶统计量方法用于滚动轴承故障特征提取进行了研究,提出了基于双谱估计的滚动轴承故障诊断方法。利用这种方法可以同时获得包含滚动轴承故障信号幅值和相位信息的双谱特征图谱。研究表明,双谱图谱可以有效地监测滚动轴承工作状态的模式,因而可以快速地识别滚动轴承不同的故障特征。

When rolling bearings are in fault,the measured vibration signals commonly are non-Guassian and non-linear with strong noise under different fault condition,the application of the higher-order statistics to extract failure features of rolling bearings is studied.A method based on bi-spectrum rolling bearing fault diagnosis is presented.With this method,the feature bi-spectrums which contain phase and magnitude of failure signals of rolling bearing can be obtained at the same time,the results show that bi-spectrum method can effectively monitor working status of the rolling bearing and can identify different faults features quickly.