基于FV-FBE的滚动轴承故障诊断研究

Fault Diagnosis of Rolling Bearing Based on FV-FBE

针对单通道信号不能全面提取旋转机械的振动信息,为了从强背景噪声中准确提取出滚动轴承的微弱故障特征,提出了一种全矢频带熵(FV-FBE)的滚动轴承故障诊断算法。该方法采用短时傅里叶变换计算频带熵(FBE),根据FBE最小原则自适应设计双通道信号的带通滤波器带宽和中心频率,对滤波后的双通道信号采用全矢Hilbert包络解调,得到全矢包络谱进行滚动轴承的故障识别。实验结果表明:FV-FBE算法可以全面准确地提取滚动轴承故障特征,优于谱峭度算法得到的全矢包络谱,抗干扰能力强。

It was not comprehensive to extract the vibration signal of rotating machinery from single channel signal.In order to extract the weak fault features of rolling bearing accurately from the strong background noise,a fault diagnosis algorithm of rolling bearing based on full vector frequency band entropy(FV-FBE)was proposed in this paper.The short time Fourier transform was used to calculate the frequency band entropy(FBE).And according to the FBE minimum principle,the bandwidth and center frequency of the dual channel signal band-pass filter were designed adaptively.The filtered dual channel signal was demodulated by the full vector Hilbert envelope,and the full vector envelope spectrum was obtained for fault diagnosis of rolling bearing.The experimental results showed that the FV-FBE algorithm could extract the fault features of rolling bearing comprehensively and accurately,which was better than the full vector envelope spectrum obtained by spectral kurtosis algorithm and had strong anti-interference ability.