基于轴承脉冲信号的早期故障检测方法研究
Research on Early Fault Detection Method Based on Bearing Pulse Signal

在本研究中，我们探讨了基于轴承脉冲信号的早期故障检测方法，重点分析了时域和频域特征以识别轴承故障。首先，通过时域分析，我们计算了信号的均方根值、峰值因子、峭度和自相关函数，从而成功捕捉到了由轴承故障引起的脉冲信号。随后，频域分析被用来进一步识别和验证故障特征频率。通过对信号进行快速傅里叶变换(FFT)，我们分析了故障前后信号的频谱变化，并利用带通滤波和谱包络技术提高了故障频率成分的检测精度。此外，引入谱峭度分析增强了故障诊断的能力，通过识别信号中的非高斯脉冲成分，准确地指出了故障发生的具体频率。这些分析结果不仅验证了所提方法的有效性，也为轴承的早期故障检测提供了强有力的技术支持。
In this study, we explored an early fault detection method based on bearing pulse signals, focusing on the analysis of time-domain and frequency-domain features to identify bearing faults. First, through time-domain analysis, we calculated the signal’s root mean square value, peak factor, kurtosis, and autocorrelation function, thus successfully capturing the pulse signals caused by bearing faults. Subsequently, frequency-domain analysis was used to further identify and verify the fault characteristic frequencies. By performing a fast Fourier transform (FFT) on the signal, we analyzed the spectral changes of the signal before and after the fault, and used band-pass filtering and spectral envelope techniques to improve the detection accuracy of the fault frequency components. In addition, the introduction of spectral kurtosis analysis enhanced the ability to diagnose faults, accurately indicating the specific frequency at which the fault occurred by identifying the non-Gaussian pulse components in the signal. These analysis results not only verified the effectiveness of the proposed method, but also provided strong technical support for early fault detection of bearings.