基于小波包分解和K最近邻算法的轴承故障诊断方法  

Bearing Fault Diagnosis Method Based on Wavelet Packet Decomposition and K Nearest Neighbor Algorithm

作　　者：朱兴统[1,2] ZHU Xing-tong(School of Automation,Guangdong University of Technology,Guangzhou Guangdong 510006,China;School of Computer,Guangdong University of Petrochemical Technology,Maoming Guangdong 525000,China)

机构地区：[1]广东工业大学自动化学院,广东广州510006 [2]广东石油化工学院计算机学院,广东茂名525000

出　　处：《装备制造技术》2020年第2期24-27,45,共5页Equipment Manufacturing Technology

基　　金：广东省自然科学基金(No.2018A030307038)。

摘　　要：轴承振动信号具有不平稳和不规则性,难以通过振动信号分析直接进行故障诊断,提出一种基于小波包分解和K最近邻算法的轴承故障诊断方法。首先利用小波包分解轴承原始振动信号,接着对分解得到的频带信号计算样本熵值,将其构建特征向量,最后利用K最近邻算法进行轴承故障诊断。并采用美国CWRU轴承数据集进行仿真实验,故障诊断效果良好,准确率为95%。Because the vibration signal of bearing is unstable and irregular,it is difficult to diagnose the fault directly by analyzing the vibration signal.In this paper,a bearing fault diagnosis method based on wavelet packet decomposition and K nearest neighbor algorithm is proposed.Firstly,the original vibration signal of bearing is decomposed by wavelet packet.Then,the sample entropy is calculated for the decomposed band signal,and these sample entropy values are taken as the eigenvector.Finally,K nearest neighbor algorithm is used for bearing fault diagnosis.The CWRU bearing data set is used for simulation experiment.The experimental results show that the fault diagnosis method is effective and its accuracy is 95%.

关 键 词：轴承 故障诊断 K最近邻算法 小波包分解 

分 类 号：TH133.33[机械工程—机械制造及自动化]