基于频率域峭度谱迭代阈值的多尺度形态学轴承故障诊断方法  

A Multi-scale Morphological Bearing Fault Diagnosis Method Based on the Iterative Threshold of the Kurtosis Spectrum in the Frequency Domain

作　　者：钱贾伟 成梁 李梦婕 QIAN Jia-wei;CHENG Liang;LI Meng-jie(School of Automation,Jiangsu University of Science and Technology,Zhenjiang 212000,China)

机构地区：[1]江苏科技大学自动化学院,江苏镇江212000

出　　处：《山东工业技术》2024年第2期111-119,共9页Journal of Shandong Industrial Technology

摘　　要：在复杂恶劣的环境中,存在滚动轴承故障信息难以准确提取的问题。因此,引用多尺度数学形态学对滚动轴承故障诊断开展研究。由于小尺度形态学滤波能较好的保存信号的细节特征,大尺度形态学滤波能有效抑制噪声,因此为了能更好的兼顾噪声抑制和故障特征信息保存,采用迭代阈值的方法选取尺度范围,运用频率域峭度谱方法计算出阈值,然后通过迭代自适应获取最佳尺度区间。多尺度形态学信号重构加权方法引用加权多尺度形态梯度算法,该方法能保证小尺度具有较小的权重,大尺度具有大权重。通过仿真和实验表明:多尺度形态学有效检测出滚动轴承故障信号,深度挖掘滚动轴承故障特征信息。In complex and harsh environment,it is difficult to extract the fault information of rolling bearings accurately.Therefore,multi-scale mathematical morphology is used to study the fault diagnosis of rolling bearings.Since small-scale morphological filtering can better preserve signal details,and large-scale morphological filtering can effectively suppress noise,in order to better balance noise suppression and fault feature information preservation,the iterative threshold method is used to select the scale range,the frequency domain kurtosis method is used to calculate the threshold value,and then the optimal scale interval is obtained through iterative self-adaptation.The multi-scale morphological signal reconstruction weighting method uses the weighted multi-scale morphological gradient algorithm,which can ensure that the small scale has a small weight and the large scale has a large weight.The simulation and experiment show that the multi-scale morphology can effectively detect the fault signal of the rolling bearing,and the fault characteristic information of the rolling bearing can be deeply mined.

关 键 词：多尺度数学形态学 迭代阈值 频率域峭度谱 加权多尺度形态梯度算法 

分 类 号：TH17[机械工程—机械制造及自动化]