自适应双阻尼小波字典的轴承复合故障诊断方法  

Bearing compound fault diagnosis method with adaptive dual-damped wavelet dictionary

作　　者：胡俊锋 赵丽娟 严雪竹 张龙 HU Junfeng;ZHAO Lijuan;YAN Xuezhu;ZHANG Long(School of Rail Transportion,Jiangxi V&T College of Communication,Nanchang 330013,China;College of Mechatronic&Vehicle Engineering,East China Jiaotong University,Nanchang 330013,China)

机构地区：[1]江西交通职业技术学院轨道交通学院,南昌330013 [2]华东交通大学机电与车辆工程学院,南昌330013

出　　处：《振动与冲击》2025年第7期239-246,共8页Journal of Vibration and Shock

基　　金：国家自然科学基金(51665013);中国国家铁路集团有限公司科技研究开发计划(N2023J042);江西省自然科学基金重点项目(20224ACB204017)。

摘　　要：针对强背景噪声下难以准确提取出轴承复合故障中各故障类型有效特征的问题,提出一种基于最大相关峭度解卷积(maximum correlated kurtosis deconvolution,MCKD)和稀疏表征的轴承复合故障诊断方法。该方法首先通过MCKD算法实现复合故障的分离,并达到初步增强故障冲击特征的效果;然后进行稀疏表征字典设计先验知识分析,构造与真实故障脉冲响应更加匹配的双阻尼非对称小波参数字典,结合正交匹配追踪算法,稀疏重构出各故障特征;最后对重构分量做包络谱分析,提取轴承故障特征频率。考虑到MCKD算法和非对称小波中的参数选取决定着最终的特征提取效果,使用鲸鱼优化算法实现参数自动优化选取。仿真数据和试验台数据分析结果表明,所提出的方法可有效提取出轴承复合故障中的各类故障成分,且相比常用的单阻尼Laplace小波字典具有一定的优越性。Here,aiming at the problem of difficulty in correctly extracting effective features of various fault types in bearing compound faults under strong background noise,a bearing compound fault diagnosis method based on maximum correlation kurtosis deconvolution(MCKD)and sparse representation was proposed.Firstly,this method could use MCKD algorithm to realize separation of compound faults and reach the effect of preliminarily enhancing fault impulse features.Then,a prior knowledge analysis of sparse representation dictionary design was performed to construct a dual-damped asymmetric wavelet parameter dictionary better matching real fault impulse responses.Combined with orthogonal matching pursuit algorithm,various fault features were sparsely reconstructed.Finally,envelope spectrum analysis was performed for the reconstructed components to extract feature frequencies of bearing faults.Considering parametric selection in MCKD algorithm and asymmetric wavelet determining the final feature extraction effect,the whale optimization algorithm was used to realize automatic and optimal selection of parameters.The analysis results of simulation data and test platform data showed that the proposed method can effectively extract various fault components in bearing compound faults,and have certain superiority over the commonly used single-damped Laplace wavelet dictionary.

关 键 词：复合故障 最大相关峭度解卷积(MCKD)算法 双阻尼非对称小波 稀疏分解 特征提取 

分 类 号：TH133.3[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]