Automatic diagnostic of transmission lines based on ultraviolet inspection

Nguyen, P.; Dudkin, Sergey Alexandrovich; Kong, Chenzai

doi:10.1051/e3sconf/201914007008

Cited by 3 publications

(1 citation statement)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Corona discharge produces an intensive ultraviolet (UV) radiation and can be detected using special UV cameras. A set of works consider some aspects of corona discharge inspection in UV spectrum [1][2][3][4]. These works reveal the following aspects: the usage of specific cameras for obtaining UV images; collecting image datasets of UV images correspondingly; performing machine learning methods including deep learning for corona discharge detection.…”

Section: Introductionmentioning

confidence: 99%

Autonomous UAV trajectory planning for corona discharge in inspection electric power lines in ultraviolet spectrum

Uzdiaev

Astapova²,

Letenkov³

2023

2nd International Conference on Computer Applications for Management and Sustainable Development of Production and Industry (CM

View full text Add to dashboard Cite

Corona discharges that occur on the conductive elements of power lines are serious problems that can lead to failures in the power system. Corona discharge produces ultraviolet (UV) radiation, which can be detected using special UV sensors. One way to automate inspection of the corona discharge is using of autonomous unmanned aerial vehicles (UAVs) equipped with UV sensor. At the same time, the trajectory of the autonomous UAV flight should be built taking into account the spatial and geometric features of the inspected power lines, as well as the requirements for the gathered data. This work describes a new method for autonomous UAV trajectory planning. The trajectory is planned in accordance with the spatial and geometric characteristics of the inspected powerline, peculiarities of the powerline damage type and in accordance with the quality requirements for the gathered data (representativeness, sample size, unification of the gathering procedure). The trajectory planning method and UV image gathering were simulated in Blender software. As a result, by the fully automated data gathering process, we collected a representative dataset of 200 images in UV spectrum for one lattice tower and two simulated discharges with 326 manually segmented corona discharge areas.

show abstract

Section: Introductionmentioning

confidence: 99%

Autonomous UAV trajectory planning for corona discharge in inspection electric power lines in ultraviolet spectrum

Uzdiaev

Astapova²,

Letenkov³

2023

2nd International Conference on Computer Applications for Management and Sustainable Development of Production and Industry (CM

View full text Add to dashboard Cite

show abstract

Localization of the Overhead Power Transmission Line Insulators Defects Using Computer Vision Algorithms

Khalyasmaa,

Zinovieva,

Mazunina

et al. 2023

2023 Belarusian-Ural-Siberian Smart Energy Conference (BUSSEC)

View full text Add to dashboard Cite

Обзор Подходов К Детектированию Дефектов Элементов ЛЭП На Изображениях В Инфракрасном, Ультрафиолетовом И Видимом Спектрах

Астапова,

Лебедев

2020

Моделирование

View full text Add to dashboard Cite

В работе представлен обзор современных методов мониторинга состояния элементов конструкции линий электропередач (ЛЭП) посредством обработки изображений в инфракрасном, ультрафиолетовом и видимом спектрах. Рассмотрены методы распознавания основных элементов конструкции ЛЭП и детектирования наиболее характерных для них дефектов, основанные на определении отличительных признаков (цвет, форма, границы, градиент яркости и текстура). В качестве основных элементов ЛЭП были рассмотрены изоляторы, провода, опоры и арматура. Анализ эффективности рассмотренных методов и подходов проводился на основе сравнения представленных в источниках метрик: значений доли верных распознаваний (accuracy), точности (precision) и полноты (recall). Особый интерес представляет анализ методов мониторинга элементов конструкции ЛЭП на основе изображений, полученных не только в видимом, но также в ультрафиолетовом и инфракрасном спектрах. Методы, предназначенные для обработки изображений в видимом спектре, основываются на алгоритмах глубокого и машинного обучения. Ультрафиолетовый спектр (УФ) используется для выявления коронных разрядов на проводах и изоляторах. Съемка в инфракрасном спектре (ИК) позволяет выявить дефекты элементов ЛЭП, которые не могут быть детектированы на изображениях в видимом спектре, например, горячие точки (hotspot). В результате проведенного анализа были рассмотрены методы детектирования дефектов ЛЭП. Методы с наибольшей эффективностью для видимого спектра: GVN, HOG + SVM, SSD, Grab cut, cascading CNN, LBP-HF + SVM, DMNN, VGG - 19, LBP + ULBP, YOLO v3, DELM + LRF, SVM, Faster R - CNN, CNN, стереозрение + PLAMEC. Методом детектирования с наибольшей эффективностью для ИК-спектра является «оцу + пороговая обработка», а для УФ-спектра метод – SVR. The paper presents an overview of modern methods for monitoring of the state of structural elements of power transmission lines (PTL) by processing images in the infrared, ultraviolet and visible spectra. Methods for recognizing of the main structural elements of power transmission lines and detecting the most characteristic defects for them, based on the determination of distinctive structural features (color, shape, borders, brightness gradient and texture), are considered. Insulators, wires, supports and fittings are considered as the main elements of power transmission lines. The analysis of the efficiency of the considered methods and approaches was performed based on the comparison of the metrics presented in the source data: values of the proportion of correct recognitions (accuracy), accuracy (precision) and recall (recall). Particularly relevant is the analysis of methods for monitoring structural elements of power transmission lines based on images obtained not only in the visible, but also in the ultraviolet and infrared spectra. Methods for image processing in the visible spectrum are based on deep and machine learning algorithms. The ultraviolet spectrum (UV) is used to detect corona discharges on wires and insulators. Imaging in the infrared spectrum (IR) enables to identify defects in power transmission lines that cannot be revealed in images in the visible spectrum, for example, hotspots. As a result of the analysis, the methods for detecting power line defects with the highest efficiency for the visible spectrum were considered: GVN, HOG + SVM, SSD, Grab cut, cascading CNN, LBP-HF + SVM, DMNN, VGG-19, LBP + ULBP, YOLO v3, DELM + LRF, SVM, Faster R-CNN, CNN, stereo vision + PLAMEC. The detection method with the highest efficiency for the IR spectrum is "Otsu + Threshold Processing", and the SVR method shows the highest efficiency for the UV spectrum.

show abstract

Automatic diagnostic of transmission lines based on ultraviolet inspection

Cited by 3 publications

References 1 publication

Autonomous UAV trajectory planning for corona discharge in inspection electric power lines in ultraviolet spectrum

Autonomous UAV trajectory planning for corona discharge in inspection electric power lines in ultraviolet spectrum

Localization of the Overhead Power Transmission Line Insulators Defects Using Computer Vision Algorithms

Обзор Подходов К Детектированию Дефектов Элементов ЛЭП На Изображениях В Инфракрасном, Ультрафиолетовом И Видимом Спектрах

Contact Info

Product

Resources

About