Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model

The surveillance and inspection of power line insulators, which act as essential components for the connection and insulation of power lines, play pivot roles in the daily maintenance of the power grid system since the failure of power line insulators could cause abrupt power cuts and accidents. Vis...

Full description

Bibliographic Details
Main Author: Zhiyong Dai
Format: Article
Language:English
Published: Elsevier 2022-11-01
Series:Energy Reports
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2352484722019205
_version_ 1828011021188464640
author Zhiyong Dai
author_facet Zhiyong Dai
author_sort Zhiyong Dai
collection DOAJ
description The surveillance and inspection of power line insulators, which act as essential components for the connection and insulation of power lines, play pivot roles in the daily maintenance of the power grid system since the failure of power line insulators could cause abrupt power cuts and accidents. Vision-assisted Unmanned Aerial Vehicle (UAV) technology has recently become an efficient and economical solution for automatic insulator fault inspection and demonstrated considerable accuracy with deep learning-based detection methods. However, current detection methods cannot predict and handle the uncertainty of insulator detection, and their capabilities are limited. This paper presents a novel deep learning-based methodology, namely YOLOD, to address the uncertainty issue by applying a Gaussian prior to the detection heads of YOLOX, the current state-of-the-art model of compact object detectors, for both bounding box regression and corresponding uncertainty estimation. Then, the estimated uncertainty scores are utilized to refine the bounding box prediction and further improve the robustness of the detection. Finally, in the comprehensive experiments, the proposed YOLOD model outperforms other benchmark models on a public insulator dataset and achieves the highest average precision (73.9%), which is 2.1% higher than that of YOLOX. Thus, the effectiveness and superiority of the proposed method for robust insulator defect inspection are validated.
first_indexed 2024-04-10T09:08:52Z
format Article
id doaj.art-80946d6f255a43f88fd9cdf3d790c08f
institution Directory Open Access Journal
issn 2352-4847
language English
last_indexed 2024-04-10T09:08:52Z
publishDate 2022-11-01
publisher Elsevier
record_format Article
series Energy Reports
spelling doaj.art-80946d6f255a43f88fd9cdf3d790c08f2023-02-21T05:13:56ZengElsevierEnergy Reports2352-48472022-11-0181280912821Uncertainty-aware accurate insulator fault detection based on an improved YOLOX modelZhiyong Dai0DeepBlue Academy of Sciences, Shanghai, 200240, ChinaThe surveillance and inspection of power line insulators, which act as essential components for the connection and insulation of power lines, play pivot roles in the daily maintenance of the power grid system since the failure of power line insulators could cause abrupt power cuts and accidents. Vision-assisted Unmanned Aerial Vehicle (UAV) technology has recently become an efficient and economical solution for automatic insulator fault inspection and demonstrated considerable accuracy with deep learning-based detection methods. However, current detection methods cannot predict and handle the uncertainty of insulator detection, and their capabilities are limited. This paper presents a novel deep learning-based methodology, namely YOLOD, to address the uncertainty issue by applying a Gaussian prior to the detection heads of YOLOX, the current state-of-the-art model of compact object detectors, for both bounding box regression and corresponding uncertainty estimation. Then, the estimated uncertainty scores are utilized to refine the bounding box prediction and further improve the robustness of the detection. Finally, in the comprehensive experiments, the proposed YOLOD model outperforms other benchmark models on a public insulator dataset and achieves the highest average precision (73.9%), which is 2.1% higher than that of YOLOX. Thus, the effectiveness and superiority of the proposed method for robust insulator defect inspection are validated.http://www.sciencedirect.com/science/article/pii/S2352484722019205Fault detectionPower line insulator inspectionDeep learningAerial image
spellingShingle Zhiyong Dai
Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
Energy Reports
Fault detection
Power line insulator inspection
Deep learning
Aerial image
title Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
title_full Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
title_fullStr Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
title_full_unstemmed Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
title_short Uncertainty-aware accurate insulator fault detection based on an improved YOLOX model
title_sort uncertainty aware accurate insulator fault detection based on an improved yolox model
topic Fault detection
Power line insulator inspection
Deep learning
Aerial image
url http://www.sciencedirect.com/science/article/pii/S2352484722019205
work_keys_str_mv AT zhiyongdai uncertaintyawareaccurateinsulatorfaultdetectionbasedonanimprovedyoloxmodel