Research On Vision Detection And Location Method Of Intelligent Inspection Robot

As an important part of power transmission,overhead high-voltage transmission lines have a large load and a long distance,making a great contribution to people’s happy life.However,due to long-term exposure to the external environment,overhead high-voltage transmission lines often suffer from corrosion,icing and other problems.Therefore,in order to ensure the safe and stable operation of high-voltage transmission lines,it is necessary to carry out regular inspection.The current inspection method has such defects as high cost and large blind area,so it becomes an urgent problem to find an advanced inspection method.With the progress of science and technology,the emergence of robots has ushered in the dawn of the high-voltage transmission line inspection problem.By using the high voltage line patrol robot to walk on the line,it can judge whether there is a fault in the line according to the picture transmitted from the on-board camera.However,some accessories on the high-voltage line will hinder the robot’s walking,so in order to ensure that the patrol robot can successfully cross the obstacles,it is necessary to first detect the types of obstacles and obtain the coordinate information of the relative robots,so as to provide the patrol robot with different signals to cross these obstacles.According to the structure characteristics of 500kV high-voltage line and the specific requirements of the task,this paper proposes a visual detection and location method for highvoltage transmission line obstacles.Firstly,this paper introduces the prototype and system composition of the inspection robot used in this project,and also introduces the working environment of the robot,focusing on the analysis of the structural characteristics of counterweight,suspension clamp and tension clamp,and completes the design of obstacle detection and positioning algorithm respectively.Among them,accurate and rapid detection and positioning of obstacle targets is the premise of obstacle surmounting for the patrol robot.The robot can choose the obstacle surmounting scheme according to the types of obstacles,and the real-time location information of obstacles can provide data support for the autonomous obstacle surmounting of the robot.This paper makes a detailed design and analysis of the target detection and location algorithm.The algorithm mainly includes two parts:obstacle detection and identification and depth information acquisition.Among them,YOLOV3 network is used to detect and identify obstacles,and then linear regression model is used to calculate the depth information of obstacles.Based on the designed model,the data set is trained by hand,and the experiment and analysis are carried out on the simulation work line set up in the laboratory.The results show that the method is applied to the detection and location of obstacles on the high-voltage line,and has achieved good results.