Point cloud acquisition and dissimilar weld seam localization of vaporizer finned tubes based on real-time robot pose and laser vision

Vaporizers have a wide range of applications in industry, and the welding of finned tubes is an important step in the manufacturing of vaporizers. The obstruction of U-shaped tubes poses a huge challenge to the detection of weld seams by visual sensors and robot welding, making it difficult to achieve automatic welding of finned tubes. Therefore, this paper designed an automatic robot welding system for finned tubes based on laser vision sensors, and innovatively proposed a point cloud construction method for dissimilar complex weld seams, which can achieve fast and high-precision collection of finned tube point clouds. Furthermore, a finned tube point cloud processing method was proposed, which can complete the calculation of circular weld seam parameters by extracting the fin end face and tubes. Finally, precise positioning of finned tube weld seams was achieved. Experimental results showed that the welding seam positioning accuracy of the method proposed in this paper is better than 0.22 mm, which meets the welding requirements of finned tubes.