| With the progress of the times and the advancement of science and technology,people’s activities of production and living are more and more increasingly dependent on electricity and the dependence is becoming higher.The society’s ability to withstand power outage,especially large-scale outage,is getting worse.Once a large-scale power outage occurs,it will generate a serious threat to public order and social security.Therefore,the stability of the operation and maintenance of the distribution network is particularly important.However,the traditional maintenance work is full of high intensity working and high safety risk in the power supply,and the problem of safety of maintenance operation is very prominent.It is of profound practical significance to study live power maintenance operation robots for distribution network to replace traditional manual operations and solve maintenance problems on distribution network.In the research of live power lines maintaining robots for distribution network,we focuses on the telepresence teleoperation technology in this thesis.We combine immersive stereo image feedback with force feedback technology to improve the telepresence of robot teleoperation.The main research carried out and results achieved are summarized as follows:1.Development of teleoperation system for live power lines maintaining robot.Research on the basic content of the 6-DOF kinematics forward solution and inverse solution,configuration method and so on are conducted for robot on the slave side.The telepresence teleoperation control system is designed including force feedback instrument on the master side,communication layer,simulation interactive system.Particularly,an immersive visual image feedback technology was developed,which used two high-definition cameras to simulate the interpupillary distance and angle of view of the human eye,and rendered the acquired binocular camera images in an immersive virtual reality helmet(HMD)to allow the operator to observe by a stereo-vision way,thereby enhancing the accuracy of robot remote operation.In this way,the accuracy of robot teleoperation is enhanced.2.A new telepresence teleoperation force feedback method combining graphic force feedback and direct force feedback is proposed.Direct force feedback,that is,by feeding back the force sensor data on the slave side of the teleoperation to the force feedback instrument on the master side,providing the operator with direct feedback of the state of the force at the end of the robots in reality and in the simulation environment.Graphic force feedback,that is,in the rendering of stereo vision,three pairs of graphic progress bars that can change their length according to the force information from the robot end tool.Within different force ranges,it will display different RGB colors(Green-Orange-Red),provides the operator with a graphical display of force feedback.In addition,in order to avoid the problem of image occlusion when operating,the position of the robot tool end is tracked synchronously to render the graphic force feedback pattern near around it.3.Research on the visual recognition and model reconstruction technology of live power maintaining targets.Based on the sparse disparity and contour operator matching method,a method is proposed for live lines’ recognition and reconstruction of the distribution network.This method includes image preprocessing,target recognition and image extraction,binocular stereo measurement,generating depth point cloud,and finally reconstruction of the 3D model of the target.At the same time,for objects with fixed shapes and features such as lightning arresters,the effectiveness is verified of applying YOLOv4 deep convolutional neural network to recognize them. |
PDF Full Text Request