基于改进特征模态分解和谱峭度的滚动轴承故障诊断  

Rolling Bearing Fault Diagnosis Based on Improved FMD and SK

作　　者：张磊 陈学军[2] 马霖 刘烽 杨康 ZHANG Lei;CHEN Xuejun;MA Lin;LIU Feng;YANG Kang(School of Mechanical and Electrical Engineering,Fujian Agriculture and Forestry University,Fuzhou 350108,China;Key Laboratory of New Energy Equipment Testing at Putian University in Fujian Province,Putian 351100,China;School of Mechanical Engineering and Automation,Fuzhou University,Fuzhou,350116,China)

机构地区：[1]福建农林大学机电工程学院,福建福州350108 [2]莆田学院新能源装备检测福建省高校重点实验室,福建莆田351100 [3]福州大学机械工程及自动化学院,福建福州350116

出　　处：《贵州大学学报（自然科学版）》2024年第5期32-40,94,共10页Journal of Guizhou University:Natural Sciences

基　　金：福建省自然科学基金资助项目(2022J011169)。

摘　　要：考虑到滚动轴承的故障信号通常表现出非平稳、容易受到干扰而无法被有效诊断的特点,提出一种改进特征模态分解(feature mode decomposition,FMD)和谱峭度(spectral kurtosis,SK)的滚动轴承故障特征提取方法。首先,对FMD的滤波特性进行研究,合理选择其两个重要输入参数;其次,采用FMD对滚动轴承故障信号进行处理,得到多个模态分量的分解结果,并通过计算各模态分量与原始信号的灰色关联度及互信息来选择最佳模态分量;最后,通过谱峭度对最佳模态分量进行带通滤波来凸显信号中的周期性冲击成分,从而有效实现故障特征频率的提取。通过分析仿真信号和实验数据的结果,发现提出方法可以有效提取故障信号特征频率的包络谱峰值和倍频谐波部分,且相比于其他方法,谱线特征更加明显,说明使用该方法对滚动轴承出现的故障进行诊断是可行的。Considering that the fault signals of rolling bearings usually exhibit non-stationary characteristics,are easily disturbed and cannot be effectively diagnosed,an improved feature mode decomposition(FMD)combined with spectral kurtosis(SK)is proposed to extract fault features of rolling bearings.First,we studyed the filtering characteristics of FMD to select two important input parameters reasonably;then,we used FMD to process the rolling bearing fault signal to obtain the decomposition results of multiple modal components,and calculated the relationship between each component and the original signal by the gray relational degree and mutual information to select the best modal component;finally,we combined with spectral kurtosis,bandpass filtering is performed on the best modal component to emphasize the impact component within the signal,thereby enabling extraction of the characteristic fault frequency.By analyzing the simulation signal and experimental data,It is found that the proposed method can effectively extract the peak envelope spectrum and harmonics of the fault signal characteristic frequency,and the spectral line features are more obvious compared to other methods,indicating that using this method to diagnose faults in rolling bearings is feasible.

关 键 词：特征模态分解 谱峭度 灰色关联分析 滚动轴承 故障诊断 

分 类 号：TH133.33[机械工程—机械制造及自动化]