Operation trajectory extraction method of mobile welding robot based on point cloud reconstruction

To ensure mobile welding robots complete high precision and efficiency operation in unstructured environments such as construction,petrochemical or shipbuilding,a welding trajectory extraction strategy was proposed based on point cloud reconstruction.The RGB-D camera was employed to the target detection to initial point guidance and reconstruction point cloud after changing the work scene.Then the point cloud trajectory was extracted and the robotic pose adjustment method was proposed based on point cloud segment and boundary estimation algorithm.Aiming at the point cloud accuracy problem,the accurate welding seam position was generated using the laser vision sensor.The seam tracking experimental of curve shaped welding seam was conducted,and the results showed that the strategy realized continuous smoothly welding seam tracking and trajectory error was less than 0.05mm,which met the production requirements of automation working. © 2024 CIMS. All rights reserved.