基于谱峭度与双谱的轴承故障诊断方法

针对强背景噪声情况下的轴承故障诊断问题，提出了一种基于谱峭度与双谱的轴承故障诊断方法。该方法采用谱峭度检测振动信号瞬态频带，利用最优带通滤波器抑制振动信号中的噪声得到滤波后信号。由带通滤波器的带宽确定低频矩形区域，计算落于该区域内的信号包络的局部双谱图，依据局部双谱图诊断轴承故障。仿真分析表明，强背景噪声和滚动体滑动会造成传统轴承故障诊断方法（如峭度、功率谱、包络谱）失效，而该方法能够有效地抑制噪声，更准确地诊断轴承故障，并通过6205-2RS JEM SKF轴承诊断实例，说明了该方法的有效性。

Rolling element bearing fault diagnosis based on spectral kurtosis and bi-spectrum

Aiming to deal with the bearing fault diagnosis problem in the case of strong background noise being contained in the collected vibration signals, a rolling element bearing fault diagnosis method based on spectral kurtosis and bi-spectrum was proposed. Spectral kurtosis was used to indicate the frequency band where the transient occurred and the optimal band-pass filter was adopted to remove the background noise. According to the bandwidth of the band-pass filter, a low-frequency rectangular area was determined, and bi-spectrum of the envelope of the band-pass filtered signal on this rectangular area was computed. The bearing fault was diagnosed according to the bi-spectrum graph. Analysis on simulation signals shows that strong background noise and rolling element slip can give rise to the failure of traditional diagnosis methods (e.g. kurtosis, power spectrum and envelop spectrum); the proposed method can inhibit the noise effectively and diagnose the rolling element bearing fault more accurately. A diagnosis instance of the bearing 6205-2RS JEM SKF was presented to show the effectiveness of the proposed method.