EUVIMEDEUVIMED entdecken
Crossref

A Contact Temperature Inversion Method for GIS Disconnector Based on Multispectral Fusion U-Net and LightGBM

Jinsheng Pan, Shengwen Shu, Sisheng Wu, Xin He, Zhiwen Bian, Jiangjun Ruan

IEEE Transactions on Instrumentation and Measurement · 2026 · 75 · 2511814-2511814

DOI: 10.1109/tim.2026.3699743 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Instrumentation and Measurement, 2026. Verfasst von Jinsheng Pan und 5 weiteren Autor:innen. Die Arbeit nennt 32 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Contact temperature is an important indicator for evaluating the operational status of gas-insulated switchgear (GIS). However, existing approaches for measuring contact temperature mainly rely on temperature sensors that are attached to the equipment enclosure, which have limited measurement coverage and require complex installation. To address these limitations, this article proposes a contact temperature inversion method for GIS disconnectors. This method integrates a multispectral fusion U-Net with a light gradient boosting machine (LightGBM) regression model, where the former is used for semantic segmentation. First, infrared and visible images of GIS disconnectors are acquired and preprocessed to construct a paired infrared–visible dataset. Second, a multispectral fusion U-Net semantic segmentation model is employed to accurately segment the three-phase enclosure regions of the GIS disconnector. Third, the segmented enclosure masks are mapped onto the raw radiometric temperature array to extract temperature statistical features for each phase enclosure. Last, these extracted statistical features are used as input to a LightGBM-based model to infer the contact temperature. The proposed multispectral fusion U-Net achieves a mean intersection over union (mIoU) of 97.12 % and a mean pixel accuracy (mPA) of 98.41 % in experimental results. Furthermore, the LightGBM contact temperature inversion model attains a coefficient of determination ( $R^{2}$ ), mean absolute error (MAE), and mean absolute percentage error (MAPE) of 0.9956, $0.228~^{\circ }$ C, and 0.66 %, respectively, demonstrating strong predictive performance. This work provides a reference for intelligent temperature monitoring of GIS disconnectors.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Pan, J., Shu, S., Wu, S., He, X., Bian, Z., & Ruan, J. (2026). A Contact Temperature Inversion Method for GIS Disconnector Based on Multispectral Fusion U-Net and LightGBM. IEEE Transactions on Instrumentation and Measurement, 75, 2511814-2511814. https://doi.org/10.1109/tim.2026.3699743

Vancouver

Pan J, Shu S, Wu S, He X, Bian Z, Ruan J. A Contact Temperature Inversion Method for GIS Disconnector Based on Multispectral Fusion U-Net and LightGBM. IEEE Transactions on Instrumentation and Measurement. 2026;75:2511814-2511814. doi:10.1109/tim.2026.3699743

BibTeX
@article{Pan2026,
  author = {Pan, Jinsheng and Shu, Shengwen and Wu, Sisheng and He, Xin and Bian, Zhiwen and Ruan, Jiangjun},
  title = {A Contact Temperature Inversion Method for GIS Disconnector Based on Multispectral Fusion U-Net and LightGBM},
  journal = {IEEE Transactions on Instrumentation and Measurement},
  year = {2026},
  volume = {75},
  pages = {2511814--2511814},
  doi = {10.1109/tim.2026.3699743}
}
Kostenlos starten →