基于MOMEDA与双谱分析的滚动轴承早期故障诊断

滚动轴承早期故障阶段，故障特征微弱且环境噪声干扰严重，采集数据包含大量噪声信息，传统的包络谱分析难以提取故障特征信息。双谱分析理论上可以抑制高斯噪声，但很难从强背景噪声下提取出微弱故障特征。而多点最优调整的最小熵解卷积（Multipoint Optimal Minimum Entropy Deconvolution Adjusted,MOMEDA）方法能增强信号中的冲击特征，但其效果和故障信号周期区间等参数有关。利用MOMEDA与双谱分析进行信号处理，将提取到的信号高阶谱特征作为滚动轴承早期故障分类依据。利用MOMEDA方法对采集信号进行滤波处理，提取出有冲击特征的时域信号；对特征增强的信号进行双谱分析，从高阶谱中提取故障特征。经过仿真信号分析和实际轴承故障信号验证，该方法能有效地提取出滚动轴承早期故障特征，实现故障诊断。

Early Fault Diagnosis of Rolling Bearing Based on MOMEDA and Bispectral Analysis

In the early stage of rolling bearing fault,the fault features are weak and the environmental noise disturbance is serious.The collected data contain a lot of noise information.It is difficult to extract the fault feature information by traditional envelope spectrum analysis.In theory,bispectrum analysis can suppress Gauss noise,but it is difficult to extract weak fault features from strong background noise.Multipoint optimal minimum entropy deconvolution adjusted (MOMEDA) method can enhance the impact characteristics of signals,but its effect is related to parameters such as fault signal period interval.MOMEDA and bispectrum analysis are used for signal processing.The high-order spectral features of the extracted signals are used as the basis for early fault classification of rolling bearings.The MOMEDA method is used to filter the acquisition signal to extract the time-domain signal with impulse characteristics,and the feature-enhanced signal is bispectral analyzed,and the fault features are extracted from the high-order spectrum.After simulation signal analysis and actual bearing fault signal verification,this method can effectively extract the early fault features of rolling bearings and realize fault diagnosis.