基于自适应STFT的货车滚动轴承故障诊断

Fault Diagnosis of Freight Car Rolling Element Bearings with Adaptive Short-Time Fourier Transform

带故障的铁路货车滚动轴承振动信号表现为低频平稳信号与高频的周期性冲击信号的叠加。采用以三阶B样条函数作为窗函数的自适应短时傅立叶变换(STFT)对货车滚动轴承振动信号进行时频分析和故障信息提取。与传统的固定带宽的STFT相比,自适应STFT在不同频段自适应选取窗长,大大提高了振动信号的时频分辨率。应用该方法对197726型货车滚动轴承在内圈剥离、外圈剥离两种故障状态下的振动信号做了分析,求得故障频率分别为61.32 Hz和46.36 Hz,与内外圈的理论故障频率相符,可以有效地诊断出铁路货车滚动轴承内外圈故障。

Vibration signals collected from freight car rolling element bearings with localized faults often consist of stationary components and periodic impulses. In this paper, an adaptive short-time Fourier transform is applied to the time-frequency analysis and feature enhancement of the vibration signals collected from freight car rolling element bearings with localized faults. Differing from the standard short-time Fourier transform, the adaptive transform uses cubic B-splines as the window functions and optimizes the window bandwidth along the frequency axis and thus improves the time-frequency resolution greatly. The method is applied to the analysis of vibration signals of 197726 type rolling element bearings with outer-race and inner-race faults. The obtained characteristic frequencies of defects are 46.36 Hz and 61.32Hz respectively which conform to the real values and show that the method performs effectively in fault diagnosis of freight car rolling element bearings.