基于LMS和Fast-Kurtogram的滚动轴承早期故障诊断

Early Fault Diagnosis of Rolling Bearings based on LMS and Fast-Kurtogram

针对滚动轴承早期故障特征提取困难的问题,提出一种LMS(Least Mean Square,LMS)算法降噪、FastKurtogram选频和共振解调技术相结合的滚动轴承故障诊断方法。首先对采集到的信号进行自适应降噪,减弱背景噪声的影响;然后利用谱峭度值对故障信号中瞬态成分敏感的特性,通过计算降噪后信号的快速峭度图,确定滤波器最优频带中心和带宽;最后进行共振包络解调提取出滚动轴承早期故障特征。通过仿真和实验验证分析,验证了该方法在滚动轴承早期故障诊断中的适用性和有效性。

Due to the difficulty of early fault features extraction of rolling bearings, a new fault diagnosis method for rolling bearings based on LMS algorithm noise reduction, Fast-Kurtogram frequency selection and resonance demodulation technology is proposed. First of all, the adaptive noise reduction is used to reduce the effect of the background noise. Then, based on the characteristics of spectral kurtosis, which is sensitive to the transient components of the faulty signal, the optimal band center and bandwidth of filter can be determined by plotting the Fast-Kurtogram of the denoised signal. Finally, the resonance envelope demodulation is used to extract the early fault characteristics of the rolling bearing. The feasibility and efficiency of this proposed method for the early fault diagnosis of rolling bearing have been verified by simulation and experiments.