High Voltage Insulators Condition Analysis using Convolutional Neural Network

Arailym Serikbay, Mehdi Bagheri, Amin Zollanvari

2021

Closed access Metadata (CSL-JSON)
DOI 10.1109/eeeic/icpseurope51590.2021.9584624
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

High voltage insulators are nonconductive materials that are in use to isolate the conductor from the earthed transmission tower. However, they are regularly subjected to operational environmental contamination and detrimental electrical or mechanical stress. Undesirable operational conditions can cause insulations’ failure and ultimately the power transmission line outage. Therefore, monitoring and condition analysis of the insulators are essential tasks for the utility operators as well as transmission companies. With the widespread applications of deep learning techniques in engineering, in this study, we propose Convolutional Neural Network (CNN) as the backbone data analysis method and aerial images as the primary dataset. We conducted a model selection through an exhaustive search within a limited hyperparameter space based on our computational resources. We show that the constructed CNN classifier achieved a remarkable accuracy of 83.3% in classifying a clean insulator surface from the one covered with water droplets.