VISION BASED SENSOR SYSTEMS FOR SEAM TRACKING AND CONTROL IN ARC AND HYBRID LASER-ARC WELDING

摘要

In recent decades much research has been performed in order to allow better control of the welding process, but the success has been very limited and the vast majority of the industrial structural welding work is therefore still being made manually. Due to the very complex nature of the welding processes, modelling and control of the processes are very demanding and the development of an intelligent closed-loop control system that acts in the same way as a manual welder is therefore needed. Closed-loop control of the processes requires the direct observation of characteristic parameters of the molten pool (e.g. geometry) in a local environment dominated by the very intense light emission from the welding arc. In laboratories it has proven possible to obtain excellent pictures of the molten pool using equipment based on weld pool illumination by an external light source typically in the form of a laser source having a high output power in the order of several hundreds watts. However, such systems are expensive and also for safety reasons not suitable in industrial applications, and today no suitable sensor exists on an industrial level that can perform this task properly. The present paper discusses various vision based techniques for performing seam tracking and closed-loop control of arc and hybrid laser-arc processes in an industrial environment. Preliminary results based on a cheap monochrome CCD camera solution are presented.