AN AUTO-PROGRAMMING SYSTEM OF MAG WELDING PARAMETERS FOR VISION-BASED ROBOT

The on-line auto-programming of MAG welding parameters for a vision-based robot mainly aims at improving the adaptability of a welding robot in a complex welding environment, and it is a new kind of intelligent control method of weld quality. The authors developed an experimental system of an MAG welding robot with 3-D vision according to the requirements of real-time auto-programming of welding parameters. In the system, the seam geometry parameters, i.e., root gap and root face in a single V-groove, are used as input variables, while the welding process parameters, i.e., welding current and speed, are used as output variables. The relation between input and output variables is described by a fuzzy model. Experimental results show that the system can result in increased weld quality.