高阶时频分布在滚动轴承故障诊断中的应用研究

Rolling Bearing Fault Diagnosis Based on Higher-order Time-frequency Distribution

滚动轴承故障信号是非高斯、非线性、非平稳信号,由于调制性,使用幅值谱分析方法无法找到轴承故障频率。将高阶时频分布——主要是Wigner双谱分析方法应用于滚动轴承的故障诊断,计算滚动轴承振动信号的高阶时频分布,并对其沿时间轴切片,分析了切片的幅值谱。针对不同故障轴承,对其振动信号幅值谱、高阶时频分布及其切片谱的对比分析表明,高阶时频及其切片谱性能优于幅值谱。

Vibration signals of defect rolling bearings are non-Gauss,non-linear and non-stationary. Fault characteristic frequencies can not be found via amplitude spectrum analysis because of modulation. Higher-order time-frequency distribution, mainly Wigner bispectrum is proposed in fault diagnosis of rolling bearings. Wigner bispectrum slice along time axis and its amplitude spectrum are given. Amplitude spectrum,higher-order time-frequency distribution and Wigner bispeetrum slice spectrum of different fault rolling bearing signal are calculated and analyzed; the results show that higher-order time-frequency distribution and its slice spec- trum have significant advantages over traditional amplitude spectrum.