Multilayer Positioning Strategy for Tubesheet Welding Robot Based on Point Cloud Model

Abstract

Weld seam positioning has been of great importance in enhancing welding automation levels. In this paper, a multilayer positioning strategy was proposed for automatic weld seam positioning based on the point cloud model of the tubesheet. The method consists of coarse positioning and fine positioning. In the process of coarse positioning, a feature-based maximum likelihood estimation (MLE) was proposed, whose solution was obtained by the proposed likelihood calculation method between the point cloud model and the tubesheet image. Considering that several different coordinate systems—robot coordinate system (RCS), tubesheet coordinate system (TCS) and point cloud coordinate system (PCS) exist in the tubesheet welding process, through the coordinates of multiple sets of MLE, the conversion relationships of these coordinate systems were obtained, so as to complete the coarse positioning process. In the process of fine positioning, the image of the weld seam was processed to solve the problem of low positioning accuracy in the coarse positioning process. Experiments showed that the positioning accuracy of the proposed multilayer positioning strategy can meet the welding requirements. Compared with the existing methods, the method in this paper has a higher level of welding automation.