Research On Automatic Tracking Technology Of Transmission Line Based On Unmanned Aerial Vehicles

With the development of our country's economy,the demand for electricity and energy is growing.In the coming years,China will build more high-voltage transmission lines.Transmission lines and power towers are the main components of power transmission and are exposed in the field environment for a long time.Because of geological disasters or their own aging and other factors,they are easily damaged.Therefore,power companies need to regularly inspect transmission lines.At present,the inspection methods based on manual inspection can no longer meet the development requirements of the electric power industry for advanced inspection technology,higher inspection efficiency,and lower inspection cost.UAV-based automatic transmission line tracking technology is part of the autonomous patrol inspection of UAV transmission lines.This article will elaborate on the following aspects:(1)Analysis of automatic tracking of power lines based on UAV,design and build an automatic tracking experimental platform for UAV power lines,and expounded its software structure.The experimental platform includes both drone and ground stations.The UAV transmits the image information of the aerial power line to the ground station in real time.The ground station processes the image information and identifies the direction information of the power line and feeds it back to the drone.The drone end The information on the direction of the power line fed back by the ground station end adjusts the direction of the operation so as to realize the automatic tracking of the power line by the drone.(2)Analyze the preprocessing process of aerial images,including graying,equalization,image restoration and image segmentation.The types of noise that cause image degradation are analyzed and Gaussian filters are used to recover aerial images.Three kinds of image segmentation algorithms,namely global adaptive threshold segmentation,local adaptive threshold segmentation and maximum information entropy,were compared experimentally.After analyzing the experimental results,the local adaptive threshold segmentation method was selected as the image segmentation method of this paper.(3)Extracting transmission line information from aerial images is accomplished in three steps: edge detection,power line identification,and power line fitting.The effects of Sobel algorithm,Laplacian algorithm and Canny algorithm on image edge detection are compared.The effects of Hough transform and LSD method on identifying transmission lines are compared.The identified lines are classified to fit the transmission line.(4)According to the principle of camera imaging,the aerial image of UAV is corrected,and the relative position relationship between UAV and transmission line is calculated.Based on the Kalman filter tracking method,the autonomous tracking model of UAV is established,and the simulation experiment is performed using Matlab software.(5)Based on the above selected method,the aerial image real-time processing software was designed using Visual C++ design language and OpenCV computer vision library.