Bi-TAM-Net framework: Fault diagnosis for insulated bearing based on new noise-resistant time-series framework DOI

Xingyuan Huang,

Tongguang Yang,

Dianjun Yang

et al.

Measurement Science and Technology, Journal Year: 2024, Volume and Issue: 36(1), P. 016112 - 016112

Published: Oct. 8, 2024

Abstract Insulated bearings are extensively employed in wind turbines and other applications as essential core parts of high-power frequency control motors. However, the influence turbine structure makes it difficult to define insulated bearing fault signal extraction. In order solve above challenges, Bi-TAM-Net framework is developed diagnose signals achieve accurate identification faults. Firstly, temporal information feature fusion model created by using time-series dataset data input with recursive chain linking rules direction evolution. Then self-attention mechanism introduced into designed for optimization, which can be modeled sequences arbitrary length, strengthening extraction ability proposed important information. Finally, based on same dataset, compared analyzed seven methods such advanced TAM-Net model, results show that has better superiority.

Language: Английский

Cross-domain knowledge transfer in industrial process monitoring: A survey DOI
Zheng Chai, Chunhui Zhao, Biao Huang

et al.

Journal of Process Control, Journal Year: 2025, Volume and Issue: 149, P. 103408 - 103408

Published: March 23, 2025

Language: Английский

Citations

0
Natural Modal Sketching Network: An Interpretable Approach for Bearing Impulsive Feature Extraction DOI
Zheng Yuan, Weihua Li, Guolin He

et al.

IEEE Transactions on Cybernetics, Journal Year: 2024, Volume and Issue: 55(2), P. 953 - 968

Published: Dec. 3, 2024

Impulsive feature (IF) response is an essential indicator for rolling bearing fault. However, it overwhelmed by strong noise and difficult to extract in real scenes. Although deep learning-based methods are powerful extraction, their logic extracting principles possess weak interpretability credibility. Their further implementation hampered. In this article, a natural modal sketching network (NMSNet) constructed achieve robust credible IF extraction. First, the designed as convolutional kernel of NMSNet, forward propagation interpreted sketching, including recovery weighted superposition. The derives from fault mechanism brings solid credibility NMSNet. Second, novel correction algorithm developed interpret extraction principle NMSNet theory elimination due its filter nature. Third, realizes adaptive via formulated fusion strategy training constraint. Finally, simulation experiment have been carried out verify effectiveness robustness fault-related analysis confirms knowledge acquisition which strengthens

Language: Английский

Citations

2
The STAP-Net: A new health perception and prediction framework for bearing-rotor systems under special working conditions DOI
Tongguang Yang, Dazhong Wu,

Shizheng Qiu

et al.

Reliability Engineering & System Safety, Journal Year: 2024, Volume and Issue: 254, P. 110633 - 110633

Published: Nov. 4, 2024

Language: Английский

Citations

2
Characteristic analysis and diagnosis method optimization of scroll compressor pressure pulsation signal under voltage fluctuation{fr}Optimisation des méthodes d'analyse et de diagnostic caractéristiques du signal de pulsation de pression du compresseur Scroll sous fluctuations de tension DOI
Yanjie Zhao,

Tonghe Zhang,

Yongxing Song

et al.

International Journal of Refrigeration, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 1, 2024

Citations

1
Bi-TAM-Net framework: Fault diagnosis for insulated bearing based on new noise-resistant time-series framework DOI

Xingyuan Huang,

Tongguang Yang,

Dianjun Yang

et al.

Measurement Science and Technology, Journal Year: 2024, Volume and Issue: 36(1), P. 016112 - 016112

Published: Oct. 8, 2024

Abstract Insulated bearings are extensively employed in wind turbines and other applications as essential core parts of high-power frequency control motors. However, the influence turbine structure makes it difficult to define insulated bearing fault signal extraction. In order solve above challenges, Bi-TAM-Net framework is developed diagnose signals achieve accurate identification faults. Firstly, temporal information feature fusion model created by using time-series dataset data input with recursive chain linking rules direction evolution. Then self-attention mechanism introduced into designed for optimization, which can be modeled sequences arbitrary length, strengthening extraction ability proposed important information. Finally, based on same dataset, compared analyzed seven methods such advanced TAM-Net model, results show that has better superiority.

Language: Английский

Citations

0