Path Correction of Membrane Water-Cooled Wall Climbing Robot Based on Laser Vision

Abstract

To ensure that the wall-climbing robot can accurately walk along the predetermined straight route, it is necessary to obtain the real-time offset and deflection of the wall-climbing robot to provide information feedback for the control system. A set of double crosshair laser vision systems and corresponding image processing processes are designed, which controls the wall climbing robot with the center point between the cross intersections as the origin. The obtained offset and deflection are independent of each other, which can reduce the difficulty of control. Through the image processing method, the connection points between laser lines, fins, and water-cooled pipes can be effectively obtained, and then the offset and deflection amount can be calculated according to the obtained connection points. The average deviations between the offset and deflection obtained by the algorithm and the actual measurement are 0.74 mm and 0.9° respectively.