Research On Visual Guidance Technology For High-voltage Power Transmission Line De-icing Robot

The ice coating and snow in transmission power line can cause many poweraccidents, such as power wire breaking, tower falling down, line tripping, icingflashover, communicating discontinuity, pollution flashover and so on. At the same time,the power accidents almost come with the huge economic losses and social instability.The traditional solutions in ice costing and snow are artificial de-icing or DCice-melting technology. Both methods have a common drawback that：the needs tointerrupt the normal power transmission, especially manual de-icing directly affectsthe lives and safety of staff. Therefore, with the development of a safe and novel, longrunning, efficient and cost-saving special working robot, have became the world’sresearch focus. This dissertation carried out deicing robot research with the support ofthe National Science and Technology Support Project.Firstly, the destruction of harsh natural disasters, like snow and ice, inhigh-voltage transmission line of power transmission is introduced. Then it analyzesthe current situation and development direction of the special operation robot for highvoltage transmission line at home and abroad. It mainly discusses the current researchon inspection robots and deicing robots, as well as the research emphases anddifficulties.Secondly, a three arm obstacle detouring deicing robot is developed combinedwith the special natural environment in China and mass research of inspection robotand deicing robot all the countries in the world. The details of electrical system,software system structure and control system structure are introduced.Thirdly, the establishment of three-arm deicing robot’s kinematics equation useD-H method is finally completed with an in-depth study on mechanical structure ofthe three-arm robot system. Because the obstacle, detouring is an importantperformance index; this paper describes the process of how the deicing robot strideacross the obstacle.Fourthly, a cable positioning algorithm based on texture characteristics ispresented. Because the cable is composed of a plurality of wires twisted together,the surface textures clear. To get the positions of cable, Gabor transform can extractthe interesting area from the background, and then locate wire position with Hough transform. This algorithm is more useful than remote artificial operation in grabbingthe cable of the robot’s arm stride across the obstacle.Finally, it presents an algorithm for3D reconstruction based on binocular vision.The cable positioning problems mentioned in the former chapter is to solve themechanical arm obstacle detouring and the cable local positioning. In practicalapplications, global positioning is more useful than local positioning, and it willprovide an important information basis for a stable and reliable operation of the robot.