Development of the Inspection Robots for Power Transmission Lines from CASIA

Guodong Yang1, Wenkai Chang2, En Li1, Zize Liang1

  • 1Institute of Automation, Chinese Academy of Sciences
  • 2Institute of Automation Chinese Academy of Sciences

Details

10:00 - 10:30 | Mon 25 Sep | Ballroom Foyer | MoAmPo.39

Session: Monday Posters AM

Abstract

Since it’s dangerous and time-consuming to inspect the power transmission line by workers, robot technologies have been applied in this domain. There are two main research directions in automatic inspection according to the operation mode. One is the contact mode using robot crawling on the line. The other is the non-contact mode using UAV (Unmanned Aerial Vehicle) flying along the line. For the crawling mode, detailed line status can be obtained because the robot is close to the line. It also has multiple operation abilities and good endurance due to the large load capacity. But the complexity of line structure and the finite degree of freedoms of the robot limit the robot’s ability on negotiating obstacles like insulator strings and strain clamps. In the opposite, it is good at obstacle negotiation for the flying mode, but lack of inspection accuracy and operation stability. To provide better inspection results, the above two modes must be combined. Our team has worked on this field for decades and made some achievements on both directions. We developed a bi-branchiate robot with a symmetric structure which can crawl on the line. The robot has two symmetrical arms and a wheel with a gripper at the top of each arm. The two arms can make a relative motion along the guide rail. With the wheels and grippers, the robot can roll or crawl along the power lines. When encountering obstacles, the robot can make a so called symmetric movement to cross them. We adopt a hierarchical structure and a modular approach to implement the control system, including the visual system and the motion control system. Using monocular vision, the position and width of line from the image are chosen as the image features, and an IBVS (Image Based Visual Servoing) controller is developed to fulfill the obstacle negotiation procedure. The robot has now been tested on live lines and the experiment results show its effectiveness on power line inspection. To improve the obstacle negotiation ability, we also developed a robot with hybrid operation mode. The robot is built based on a quad-rotor platform with a trumpet-shaped undercarriage. Two wheels are equipped on the top of the undercarriage, making the robot capable of landing and rolling on the line. So it combines the major advantages of both modes, such as the obstacle negotiation ability by flying over the line and the high inspection accuracy and stability by crawling on the line. By equipping a 2D LRF (Laser Range Finder), the robot can detect the orientation of the line and recognize the obstacles on it. An adaptive algorithm is designed to control the robot negotiating the obstacles and landing on the line. We have tested the robot prototype under simulated environment and the results show that the robot can fly over obstacles and land on the line automatically. Next, we will optimize its electromagnetic characteristics to satisfy the requirements under the live line environment.