Intelligent fault diagnosis method using small fault samples driven by digital data and feature enhancement

被引：0
|
作者：
Xia J. [1 ]
Huang R. [2 ,3 ]
Chen Z. [1 ,2 ]
Li J. [1 ]
Li W. [1 ,2 ]
机构：
[1] School of Mechanical & Automotive Engineering, South China University of Technology, Guangzhou
[2] Guangdong Artificial Intelligence and Digital Economy Laboratory, Guangzhou
[3] Shien-Ming Wu School of Intelligent Engineering, South China University of Technology, Guangzhou
来源：
Zhongguo Kexue Jishu Kexue/Scientia Sinica Technologica | 2023年 / 53卷 / 07期
关键词：
digital twin; generative adversarial networks; intelligent fault diagnosis; mechanical equipment; small fault samples;
D O I：
10.1360/SST-2023-0018
中图分类号：
学科分类号：
摘要：
With the rapid development of new-generation artificial intelligence technology, intelligent fault diagnosis methods have garnered considerable attention, finding various applications in diverse industry scenarios, such as aerospace, ocean engineering, and automobile industries. However, there is a lack of sufficient fault samples for effectively training an intelligent fault diagnosis model. This shortage may affect the stability and reliability of this model. Furthermore, the existing intelligent diagnosis methods for small fault samples require highly correlated measured data sets. Therefore, this paper proposes an intelligent fault diagnosis method for industrial equipment that utilizes small fault samples driven by digital data and feature enhancement. First, a virtual model of the equipment is constructed and updated using the vibration mechanism knowledge and the measured data with the health state. Subsequently, digital fault data featuring typical fault modes are generated using the virtual model. The characteristics of the digital fault data are enhanced using state-of-the-art artificial intelligence algorithms, such as generative adversarial networks. The enhanced data are then used to train an intelligent fault diagnosis model based on convolutional neural networks. Finally, the feasibility of the proposed method is verified using a case study of vehicle transmission. The results suggest that the proposed method can provide a promising solution for intelligent fault diagnosis of modern equipment in practical industrial applications under small fault samples. © 2023 Chinese Academy of Sciences. All rights reserved.
引用
收藏
页码：1202 / 1213
页数：11
相关论文
共 25 条
  • [1] Zang J Y, Liu Y F, Wang B C, Et al., Technology forecasting and roadmapping of intelligent manufacturing by 2035, J Mech Eng, 58, pp. 285-304, (2022)
  • [2] Shang C, You F., Data analytics and machine learning for smart process manufacturing: Recent advances and perspectives in the big data era, Engineering, 5, pp. 1010-1016, (2019)
  • [3] Li J, Li X, Xu Y M, Et al., Recent advances and prospects in industrial AI and applications, Acta Autom Sin, 46, pp. 2031-2044, (2020)
  • [4] Huang R Y, Li J P, Wang Z, Et al., Intelligent diagnostic and prognostic method based on multitask learning for industrial equipment, Sci Sin Tech, 52, pp. 123-137, (2022)
  • [5] Liu R, Yang B, Zio E, Et al., Artificial intelligence for fault diagnosis of rotating machinery: A review, Mech Syst Signal Processing, 108, pp. 33-47, (2018)
  • [6] Zhao Z, Li T, Wu J, Et al., Deep learning algorithms for rotating machinery intelligent diagnosis: An open source benchmark study, ISA Trans, 107, pp. 224-255, (2020)
  • [7] Fu S, Zhong S S, Lin L, Et al., Gas turbine fault diagnosis method under small sample based on transfer learning, Comp Integrated Manuf Syst, 27, pp. 3450-3461, (2021)
  • [8] Yang B, Lee C G, Lei Y, Et al., Deep partial transfer learning network: A method to selectively transfer diagnostic knowledge across related machines, Mech Syst Signal Processing, 156, (2021)
  • [9] Han T, Liu C, Yang W, Et al., A novel adversarial learning framework in deep convolutional neural network for intelligent diagnosis of mechanical faults, Knowledge-Based Syst, 165, pp. 474-487, (2019)
  • [10] Shao H D, Li W, Liu Y, Et al., Fault diagnosis of rotor-bearing system under time-varying speeds by using dual-threshold attention-embedded GAN and small samples, J Mech Eng, 59, pp. 1-9, (2023)
123