电子学报 ›› 2020, Vol. 48 ›› Issue (8): 1647-1654.DOI: 10.3969/j.issn.0372-2112.2020.08.026

• 学术论文 • 上一篇    下一篇

样本空间基于多级高维特征表示的微小故障诊断

张彩霞1,3, 王子涵2, 文成林2, 刘国文1,3, 余伟1,3   

  1. 1. 佛山科学技术学院机电工程与自动化学院, 广东佛山 528000;
    2. 杭州电子科技大学自动化学院, 浙江杭州 310018;
    3. 广东省智慧城市基础设施健康监测与评估工程技术研究中心, 广东佛山 528000
  • 收稿日期:2018-02-22 修回日期:2018-06-24 出版日期:2020-08-25
    • 作者简介:
    • 张彩霞 女,1976年生于河南平顶山,博士,教授,研究方向为多源信息融合与智能控制系统. E-mail:zh_caixia@163.com
    • 基金资助:
    • 国家自然科学基金 (No.61803087); 广东省基础与应用基础研究基金粤港澳应用教学中心项目 (No.2019KTSCX192); 佛山市核心技术攻关项目 (No.1920001001367); 佛山市科技创新项目 (No.2016AG10011)

Sample Space Based on Multi-level High Dimensional Feature Representation Micro-fault Diagnosis

ZHANG Cai-xia1,3, WANG Zi-han2, WEN Cheng-lin2, LIU Guo-wen1,3, YU Wei1,3   

  1. 1. School of Machatronic Engineering and Automation, Foshan University, Foshan, Guangdong 528000, China;
    2. Automatic School, Hangdian University, Hangzhou, Zhejiang 310018, China;
    3. Guangdong Province Smart City Infrastructure Health Monitoring and Evaluation Engineering Technology Research Center, Foshan, Guangdong 528000, China
  • Received:2018-02-22 Revised:2018-06-24 Online:2020-08-25 Published:2020-08-25
    • Supported by:
    • National Natural Science Foundation of China (No.61803087); Program of the Guangdong-Hong Kong-Macao Applied Mathematics Center ,  Basic and Applied Basic Research Foundation of Guangdong Province (No.2019KTSCX192); Foshan Core Technology Research and Development Program (No.1920001001367); Foshan Science and Technology Innovation Project (No.2016AG10011)

摘要: 传统主元分析(Principal Component Analysis,PCA)、相对主元分析等多元统计法基于阈值诊断故障,由于是原空间等价表示,并未增加任何信息量,使得微小故障难以诊断;且降维分成主元空间和残差空间,微小信息得不到充分表示.深度学习在模式识别方面有成功的应用,深度学习多层次网络对细节进行线性组合表示,但不具备可解释性,仅有训练结果无理论依据,机理分析困难.本文提出一种将主元分析思想与深度学习思想结合的故障诊断方法,在原PCA基础上先扩维再降维,使得原始空间中不能表达的信息充分表达,且具备可解释性.理论和仿真实验分析表明,本文方法能判断出传统PCA无法检测的微小故障,提高了故障检测的检出率,且具备可解释性.

 

关键词: 多级高维, 主元分析, 投影标架, 故障诊断

Abstract: Traditional principal component analysis, relative principal component analysis and other multivariate statistical methods based on threshold to do the fault diagnosis. Since multivariate statistical method is an equivalent representation of the original space, it does not add any amount of information, making it difficult to diagnose minor faults. And the original space is reduced dimensionally into the principal component space and the residual space, so that the tiny information cannot be fully expressed. Deep learning has been successfully applied in pattern recognition. However, multilevel networks of deep learning represent linear combinations of details but do not have explanatory. Only with the result of training without theoretical basis, it is difficult to carry out mechanistic analysis. This paper presents a fault diagnosis method which combines PCA thought and deep learning thought. Based on the original PCA, this paper first expands and then reduces the dimension, making the inexplicit information in the original space fully expressed and interpreted. Theoretical and simulation experiments show that this method can judge the minor faults which cannot be detected by traditional PCA, improve the detection rate of fault detection and have interpretability.

 

Key words: multilevel high-dimensional, principal component analysis, projection frame, fault diagnosis

中图分类号: