Seam Tracking in an Unstructured Environment: A Case Study with Vee-Groove Joints
This chapter discusses the algorithms developed for a seam tracking system comprising of a six-axis, articulated joint robot and a scanning beam range sensor. Operation of an adaptive robotic welding system within the unstructured environment of a general 3D seam involves, (1) controlling the torch-seam interaction for achieving the desired weld qualities, and (2) controlling the position and orientation of the torch as well as the rotation of the lookahead range sensor for tracking the seam. These functions are governed by the high-level controller and the low-level controller, respectively. The inputs to these controllers are derived from the analysis of the range images. Specifically, the range image processing stage extracts the joint features and outputs their coordinates in world space, which is then used in the generation of the seam environment model and the seam geometry model.
KeywordsFeature Point Coordinate Frame Range Image Robot Controller Feature Extraction Algorithm
Unable to display preview. Download preview PDF.