基于ACMD与K-SVD的滚动轴承微弱故障特征诊断  

Weak Fault Feature Diagnosis of Rolling Bearings Based on ACMD and K-SVD

作　　者：牛柱强 张玮 许书庆 张成城 胡鑫磊 NIU Zhuqiang;ZHANG Wei;XU Shuqing;ZHANG Chengcheng;HU Xinlei(CNNC Nuclear Power Operation Management Corporation,Jiaxing 314300,China;Suzhou Veizu Equipment Diagnosis Technology Co.,Ltd.,Suzhou 215211,China)

机构地区：[1]中核核电运行管理有限公司,浙江嘉兴314300 [2]苏州微著设备诊断技术有限公司,江苏苏州215211

出　　处：《轴承》2025年第3期97-103,共7页Bearing

基　　金：国家自然科学基金资助项目(U2267206)。

摘　　要：针对强背景噪声下滚动轴承故障特征难以提取的问题,提出了一种基于自适应啁啾模态分解(ACMD)和K-奇异值分解(K-SVD)的故障特征提取方法。采用ACMD自适应地将原始信号分解为不同的本征模态分量,提出一种新的衡量信号故障特征的信号特征因子筛选包含故障信息丰富的模态分量作为训练信号,利用KSVD字典学习针对训练信号训练字典库,结合正交匹配追踪算法对原始信号进行重构得到稀疏信号,通过进一步的包络谱分析获取故障特征频率并作出故障诊断。仿真信号和试验研究表明,基于ACMD与K-SVD的方法能够有效提取强背景噪声下的滚动轴承故障特征,确定轴承故障类型。Aimed at the problem that the fault feature of rolling bearings is difficult to extract under strong background noise,a fault feature extraction method is proposed based on adaptive chirp mode decomposition(ACMD) and Ksingular value decomposition(K-SVD).The ACMD is used to adaptively decompose the original signal into different intrinsic mode components,and a new signal characteristic factor for measuring the fault features of signal is proposed to screen the mode components containing rich fault information as training signal.A dictionary library is trained for training signal by using K-SVD dictionary learning,and the original signal is reconstructed by combining with an orthogonal matching pursuit algorithm to obtain the sparse signal.The fault feature frequency is obtained through further envelope spectral analysis,and the fault diagnosis is carried out.The simulated signal and experimental studies show that the method based on ACMD and K-SVD can effectively extract the fault features of the bearings under strong background noise and determine the fault types of the bearings.

关 键 词：滚动轴承 故障诊断 特征提取 信号处理 奇异值分解 

分 类 号：TH133.33[机械工程—机械制造及自动化]