Slippage fault diagnosis of dampers for transmission lines based on faster R-CNN and distance constraint

Abstract

The slippage fault diagnosis of dampers in aerial images is a challenging task for an automatic transmission line inspection system. In this paper, an effective and reliable slippage fault diagnosis method based on a deep learning technique and distance constraint method for aerial images is proposed. In the detection stage, the Faster Regions with Convolutional Neural Network (R-CNN) is employed. It takes the aerial images as input, detects the dampers in the image by learning the basic features of dampers, and outputs the coordinates of the dampers in a single image. In the diagnosis stage, the distance between adjacent dampers in the horizontal axis is calculated through the damper coordinates obtained in the detection stage, and the relationship between the distance and the selected threshold is the judgment basis of slippage fault diagnosis. The results show that the slippage fault of dampers can be quickly and accurately diagnosed in the aerial images by the proposed method.