Attention mechanism in intelligent fault diagnosis of machinery: A review of technique and application. (August 2022)
- Record Type:
- Journal Article
- Title:
- Attention mechanism in intelligent fault diagnosis of machinery: A review of technique and application. (August 2022)
- Main Title:
- Attention mechanism in intelligent fault diagnosis of machinery: A review of technique and application
- Authors:
- Lv, Haixin
Chen, Jinglong
Pan, Tongyang
Zhang, Tianci
Feng, Yong
Liu, Shen - Abstract:
- Abstract: Attention Mechanism has become very popular in the field of mechanical fault diagnosis in recent years and has become an important technique for scholars to study and apply. The introduction of Attention Mechanism can help models achieve efficient resource allocation, improve the remote information capture capability of models, and significantly improve the performance of models for various equipment health management tasks (fault classification, life prediction, etc.) The application of Attention Mechanism in machinery has achieved fruitful research results, but there is a lack of related reviews. In order to facilitate later scholars to quickly grasp the Attention Mechanism and select the appropriate technique, this paper reviews the relevant research and applications of Attention Mechanism in Intelligent Fault Diagnosis of Machinery. Based on the methods proposed in the collected literature, this paper classifies and analyzes them from multiple perspectives to help readers grasp the development status and trends in this field. We divide the collected technologies into three categories: Recurrent-based, Convolution-based, and Self-attention-based. We describe each attention technique and its application scenarios in detail. Finally, we summarize the advantages and disadvantages of various AM techniques, and further discuss the possible future directions of attention mechanisms in the mechanistic field. The purpose of this paper is to provide a comprehensiveAbstract: Attention Mechanism has become very popular in the field of mechanical fault diagnosis in recent years and has become an important technique for scholars to study and apply. The introduction of Attention Mechanism can help models achieve efficient resource allocation, improve the remote information capture capability of models, and significantly improve the performance of models for various equipment health management tasks (fault classification, life prediction, etc.) The application of Attention Mechanism in machinery has achieved fruitful research results, but there is a lack of related reviews. In order to facilitate later scholars to quickly grasp the Attention Mechanism and select the appropriate technique, this paper reviews the relevant research and applications of Attention Mechanism in Intelligent Fault Diagnosis of Machinery. Based on the methods proposed in the collected literature, this paper classifies and analyzes them from multiple perspectives to help readers grasp the development status and trends in this field. We divide the collected technologies into three categories: Recurrent-based, Convolution-based, and Self-attention-based. We describe each attention technique and its application scenarios in detail. Finally, we summarize the advantages and disadvantages of various AM techniques, and further discuss the possible future directions of attention mechanisms in the mechanistic field. The purpose of this paper is to provide a comprehensive reference for researchers and to help them find further research directions. … (more)
- Is Part Of:
- Measurement. Volume 199(2022)
- Journal:
- Measurement
- Issue:
- Volume 199(2022)
- Issue Display:
- Volume 199, Issue 2022 (2022)
- Year:
- 2022
- Volume:
- 199
- Issue:
- 2022
- Issue Sort Value:
- 2022-0199-2022-0000
- Page Start:
- Page End:
- Publication Date:
- 2022-08
- Subjects:
- Fault Diagnosis -- Attention mechanism -- Deep learning -- Fault classification
Weights and measures -- Periodicals
Measurement -- Periodicals
Measurement
Weights and measures
Periodicals
530.8 - Journal URLs:
- http://www.sciencedirect.com/science/journal/02632241 ↗
http://www.elsevier.com/journals ↗ - DOI:
- 10.1016/j.measurement.2022.111594 ↗
- Languages:
- English
- ISSNs:
- 0263-2241
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 5413.544700
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 22858.xml