摘要
针对轴承故障诊断中产生数据量庞大,对存储和计算能力造成巨大负担的情况,提出一种基于压缩采样与多步特征提取的轴承故障诊断新方法。首先,采用压缩感知(CS)采集轴承信号;其次,使用主成分分析(PCA)和线性判别分析(LDA)串联的多步特征提取方法PLC对未重构的采集信号进行有效特征提取;最后,通过多分类的PSO-SVM算法来训练、验证和分类轴承故障。结果表明,所提方法在正常和噪声干扰的情况下都可以保证较高分类精度的同时,减少了计算时间,测量数据量更少,减少存储需求。
In response to the large amount of data generated in bearing fault diagnosis,which places a huge burden on storage and computing power,this paper proposes a new method for bearing fault diagnosis based on compression sampling and multi-step feature extraction.First uses compressed sensing(CS)to collect bearing signals,and uses a multi-step feature extraction method in series of principal component analysis(PCA)and linear discriminant analysis(LDA)to perform effective feature extraction on unreconstructed collected signals.Finally,the multi-classified PSO-SVM algorithm is used to train,verify and classify bearing faults.The results show that the proposed method can ensure higher classification accuracy under normal and noise interference,while reducing the calculation time,the measurement data volume is less,and reduced storage requirements.
作者
张帆
陆见光
唐向红
ZHANG Fan;LU Jian-guang;TANG Xiang-hong(Key Laboratory of Advanced Manufacturing Technology,Ministry of Education,Guizhou University,Guiyang 550025,China;School of Mechanical Engineering,Guizhou University,Guiyang 550025,China;State Key Laboratory of Public Big Data,Guizhou University,Guiyang 550025,China)
出处
《组合机床与自动化加工技术》
北大核心
2021年第4期108-112,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
贵州省留学回国人员科技活动择优资助项目(2018.0002)
贵州省公共大数据重点实验室开放基金(2017BDKFJJ019)
贵州大学引进人才基金(贵大人基合字(2016)13号)。
关键词
轴承故障诊断
压缩感知
主成分分析
bearing fault diagnosis
compressed sensing
principal component analysis
