Acquisition and optimization of weld trajectory and pose information for robot welding of spatial corrugated web sheet based on laser sensing

Abstract

Robot welding provides us a new inexpensive technology to fabricate corrugated web H-beam compared to traditional mill process. The key and challenged technology to realize the intelligent welding of the spatial curve seam is to acquire the trajectory and pose information. A practically viable and robust method based on laser sensing was proposed to solve the aforementioned key problem. Firstly, robot, laser sensor, CCD camera, encoder, working station, and computer to control the robot were integrated into a system. Next, the trajectory information was acquired by laser distance ranging sensor and encoder based on triangulation theory. The filtering and smoothing algorithm was developed to accurate the web data. The posture was established based on the trajectory information. The interactive communication between the computer and robot was established to in time control the moving of the robot. Lastly, the data which were processed by the improved amplitude-limiting filtering algorithm were transmitted to the robot. The robot executed the program files in order to complete the welding process. The results show that this method has a high practicality in robot automatic welding of corrugated web H-beam.