Identificação de falhas em linhas de transmissão de energia elétrica utilizando Deep Learning

Abstract

The Electric Sector is one of the most important for society as a whole. Its operation is responsible for improving the quality of life for citizens in almost every corner of the world. Thus, we can power our refrigerators, charge our cellphones, watch television, among many other applications that electrical energy enables. This present work demonstrates the utilization of the YOLOv8 architecture for object detection in power transmission lines, aiming to prevent or detect possible failures using the PLT-AI Furnas dataset. The obtained results were satisfactory, with an average precision of 92.1.