CREST: A target-point-based wall-climbing robot capable of spatial traversal and obstacle avoidance

Abstract

The internal inspection of smooth and clean cavity structural parts such as glass components can only be carried out by negative pressure adsorption wall climbing robots. Due to the complexity of the internal structure of these structural parts, robots are required to avoid obstacles and transition between different surfaces. However, due to the large number of joints, it is very difficult to achieve these actions and the cost is high. In this paper, we propose a method to realize the robot's space motion by controlling the robot's stride length and steering angle. Firstly, the normal climbing and traversing climbing movements are designed, the key control points (target points) are found to realize these movements, and the geometric relationship of each joint angle is derived. Secondly, we analyze the relationship between stride length and joint angle, as well as the relationship between the distance to transition the wall and stride length, and further deduce the equations of space movement, and design the equations and program of iterative calculation. With it, the robot can automatically detect the distance, adjust the stride length, and automatically transition during the climbing process. Real-world experiments have confirmed that the robot can perform straight and turning motions on horizontal, vertical and inverted surfaces, transition from one wall to another, and avoid obstacles on both horizontal and vertical surfaces. This paper reduces the difficulty of the wall-climbing robot control, and enhances the movement flexibility and application potential in the narrow environment.

Keywords

• Wall climbing robot
• Variable stride control
• Autonomous
• Motion control
• Autonomous transitioning