EUVIMEDEUVIMED entdecken
Crossref

Research on Large Model-Enabled Precise Alarm Localization and Topology Visualization Technology for Heterogeneous Cloud Resources

Jing Bai, Yan Shi, Yanxu Jin

Textile & Leather Review · 25.04.2026 · 9 · 3474-3496

DOI: 10.31881/tlr.2026.3474 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Textile & Leather Review, online erschienen am 25.04.2026. Verfasst von Jing Bai und 2 weiteren Autor:innen. Die Arbeit nennt 22 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Heterogeneous clouds consist of components at different levels from various vendors, mirroring the textile manufacturing sector where resources are characterized by diversity in fabric processing technologies, dynamic production requirements, and complex business associations between digital management systems and physical machinery. Traditional alarm localization methods struggle to achieve precise fault tracing, and topology visualization technologies fail to effectively support operation and maintenance (O&M) personnel in intuitively grasping resource correlation relationships. To address these challenges, this paper proposes a Pre-trained Language Model (PLM)-enabled method for precise alarm localization and topology visualization of heterogeneous cloud resources. Firstly, a PLM-based alarm preprocessing model is constructed; secondly, an alarm localization model integrating PLM-based and graph neural networks (GNNs) is designed; finally, a knowledge graph-based topology visualization scheme is proposed to visually present heterogeneous cloud resources, alarm information, and fault correlation relationships in the form of a knowledge graph. Experimental results show that the proposed method outperforms traditional methods in terms of alarm denoising accuracy, fault localization accuracy, and localization latency. Additionally, the operational convenience and information completeness of the topology visualization interface have been recognized by O&M personnel. This research provides a new technical approach for the efficient O&M of heterogeneous cloud resources, specifically tailored to handle the complex data environments of textile industrial clusters while holding significant theoretical significance and practical application value for digitalized garment manufacturing. By optimizing cloud management, this system offers a robust framework for integrating automated weaving data and production monitoring into a unified, highperformance operational model.

Zitieren

APA

Bai, J., Shi, Y., & Jin, Y. (2026). Research on Large Model-Enabled Precise Alarm Localization and Topology Visualization Technology for Heterogeneous Cloud Resources. Textile & Leather Review, 9, 3474-3496. https://doi.org/10.31881/tlr.2026.3474

Vancouver

Bai J, Shi Y, Jin Y. Research on Large Model-Enabled Precise Alarm Localization and Topology Visualization Technology for Heterogeneous Cloud Resources. Textile & Leather Review. 2026;9:3474-3496. doi:10.31881/tlr.2026.3474

BibTeX
@article{Bai2026,
  author = {Bai, Jing and Shi, Yan and Jin, Yanxu},
  title = {Research on Large Model-Enabled Precise Alarm Localization and Topology Visualization Technology for Heterogeneous Cloud Resources},
  journal = {Textile & Leather Review},
  year = {2026},
  volume = {9},
  pages = {3474--3496},
  doi = {10.31881/tlr.2026.3474}
}
Kostenlos starten →