Research And Application Of Transmission Line Defect Recognition Method Based On Unmanned Aerial Vehicle Automatic Inspection Technology

Transmission line equipment has been operating in the complex and changeable open air environment for a long time,and there are various defects and hidden dangers that affect the safe operation of transmission line.In order to ensure the safe and reliable operation of the transmission line,it is necessary for the power workers to inspect the transmission line regularly,so as to detect and eliminate the defects affecting the normal operation of the line as soon as possible.However,due to its low efficiency and high risk,the traditional manual inspection method can not meet the requirements of lean management.The promotion of UAV autopilot technology in transmission line inspection not only greatly reduces the workload and pressure of operation and maintenance personnel,but also improves the efficiency and quality of overhead transmission line inspection to a certain extent.However,the efficient inspection method brings a lot of image data,which need to be manually checked and analyzed by the operation and maintenance personnel to confirm the operation status and health of the line,and then determine the subsequent control level and inspection plan.In addition,the manual sorting and analysis of images often need a lot of manpower and time,and different analysis results will be produced due to the different evaluation criteria of operation and maintenance personnel.Therefore,it is of great engineering significance and theoretical value to study the transmission line defect identification method combined with UAV automatic inspection technology to overcome the tedious task of data sorting and defect analysis of multi-rotor UAV in power transmission specialty.In view of the shortcomings of the current manual analysis methods for machine patrol images,this paper firstly ensure the accurate function location of the defect according to the coordinate information and autopilot route of the aircraft patrol image;Then,used the component detection model and defect detection model which are trained by the deep learning framework of the Paddlepaddle and the cascade target detection structure to detect three typical defects of glass insulator porcelain bottle self explosion,composite insulator grading ring loose and hardware bolt cotter pin missing,at the same time,used the method that combine infrared image temperature extraction algorithm and component detection network to realize the detection of infrared image heating defects;Finally,RPA robot is used to input the output test results of the model into the production system,thus realizing the automatic processing of the whole process from the original machine inspection image to the defect management system,which greatly improves the processing efficiency of the machine inspection data.This paper mainly introduces the post processing and analysis method of UAV automatic driving inspection image data,and develops the program of batch automatic renaming of aircraft patrol photos and image defect recognition system.In the application practice of the inspection of the overhead transmission lines in Foshan area,the method can greatly reduce the workload of the operation and maintenance personnel and greatly improve the efficiency of the defect identification of the transmission lines,which proves the effectiveness and rationality of the method.