Research On Identification Method Of Broken Insulator String In Transmission Line

As an important part of the power network operation,insulators play the role of power line suspension and electrical isolation.According to statistics from the power sector,accidents caused by insulators have become the most common accidents in power operation nowadays.Therefore,it is necessary to check and maintain the insulators in time.With the development of technology,shooting of transmission lines with the help of patrol drones has become an important way of daily inspection.However,in the face of a large number of pictures returned by the drone,the inspectors can only use the manual observation method to find the defects and problems in the diagram,which leads to the final inspection results greatly affected by people’s subjective experience and fatigue degree and so on.This thesis studies the positioning problem of the shortage of insulators in the image and the main result is as follows:Firstly,aiming at the problems of low contrast,dark edge and including noise in aerial insulator image,this paper use histogram equalization to enhance the contrast of grayscale image,and then the pixel value of image edge is improved by using gradient sharpening method.Through the above steps,the difference between the pixels of the gradient image can be improved,which is beneficial to the subsequent edge detection.Secondly,for the problem that the insulator edge is difficult to be extracted completely,this paper study and analyze respectively the edge detection methods based on binary segmentation and gradient calculation.Through the analysis,it is found that the method based on gradient calculation will calculate more gray level information of the image when calculating the edge,so as to effectively extract the complete edge of the insulator,and then verify the result through experiments.Since the edge is still broken after the edge image is obtained,the edge of the image is connected by morphological operation to further optimize the edge image quality.Thirdly,about finding the position of missing insulators,this paper propose a method for identifying and positioning missing insulators based onimage contours.The method performs ellipse fitting and screening on each contour,so that the selected ellipse is located on one side of the insulator.The matching between the contours is then performed by fitting the slope and length of the elliptical center line between the different contours.Finally,the real insulator is found on the contour that matches successfully,and the position of the missing insulator is located according to its position and size,and marked in the diagram.Finally,this paper design and develop a program for detecting and locating missing insulators,and then the program is verified by experiments.The results show that the program can accurately locate the position of the missing insulator and mark it in the diagram,which has certain correctness and practicability.