Fault Diagnosis of Drilling Equipment Under Complex Conditions Based on Domain-Adversarial Transfer Learning
IEEE Internet of Things Journal · 15.05.2026 · 13 (10) · 22359-22371
DOI: 10.1109/jiot.2026.3669362 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Internet of Things Journal, erschienen am 15.05.2026. Verfasst von Lianghuai Tong und 8 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Gefördert von: State Administration for Market Regulation of Science and Technology Project, Quzhou Municipal Science and Technology Bureau Project, Deanship of Scientific Research at King Abdulaziz University, Jeddah, Saudi Arabia.
Abstract (Originalsprache)
Intelligent fault diagnosis for drilling equipment is often compromised by nonstationary noise and domain discrepancies. To address this, we propose TWE-domain-adversarial neural networks (DANNs), a novel framework integrating wavelet threshold denoising (WTD), empirical mode decomposition (EMD), and DANN. The model features a hybrid filtering architecture to suppress heterogeneous noise and employs adversarial adaptation to align cross-domain feature distributions, ensuring robust performance under complex operating conditions. Experiments on real-world drilling datasets demonstrate the effectiveness of the proposed method. On the hydraulic source-domain dataset, under the most severe composite noise conditions, TWE-DANN enables the model to reach an $F1$ -score of 98.64%. On the source (hydraulic) dataset, the model achieves an $F1$ -score of 91.32%. In the cross-domain adaptation task from hydraulic to pneumatic equipment, TWE-DANN attains an $F1$ -score of 86.98% after only five fine-tuning epochs while outperforming baseline transfer methods by 3.71%.
Förderung
- State Administration for Market Regulation of Science and Technology Project
- Quzhou Municipal Science and Technology Bureau Project
- Deanship of Scientific Research at King Abdulaziz University, Jeddah, Saudi Arabia
Zitieren
APA
Tong, L., Xu, X., Zhuge, J., Gao, B., Chen, X., Peng, X., Xu, Z., Al-Otaibi, Y. D., & Fang, K. (2026). Fault Diagnosis of Drilling Equipment Under Complex Conditions Based on Domain-Adversarial Transfer Learning. IEEE Internet of Things Journal, 13(10), 22359-22371. https://doi.org/10.1109/jiot.2026.3669362
Vancouver
Tong L, Xu X, Zhuge J, Gao B, Chen X, Peng X, et al. Fault Diagnosis of Drilling Equipment Under Complex Conditions Based on Domain-Adversarial Transfer Learning. IEEE Internet of Things Journal. 2026;13(10):22359-22371. doi:10.1109/jiot.2026.3669362
BibTeX
@article{Tong2026,
author = {Tong, Lianghuai and Xu, Xiaojie and Zhuge, Jinbing and Gao, Bo and Chen, Xing and Peng, Xueyuan and Xu, Zhongchen and Al-Otaibi, Yasser D. and Fang, Kai},
title = {Fault Diagnosis of Drilling Equipment Under Complex Conditions Based on Domain-Adversarial Transfer Learning},
journal = {IEEE Internet of Things Journal},
year = {2026},
volume = {13},
number = {10},
pages = {22359--22371},
doi = {10.1109/jiot.2026.3669362}
}
EUVIMED