Application of fuzzy control method in gas metal arc welding

In the short circuit transfer welding processes, a frequent change of arc length often causes dissatisfactory bead formation, but the control for arc length is nonlinear and complicated as it is difficult to derive a mathematical model of the controller. Thus, conventional controllers such as constant voltage power supply with constant wire feed speed cannot keep the arc length stable. This work presents a fuzzy controller using Mamdani method to stabilize the arc in gas metal arc welding (GMAW), where the deviation of arc voltage and its change rate serve as the input, and the large current at the initial arcing stage is the output. The controller is simple and easy to implement, because it has only one control object (i.e., welding current) in the system. To assess the validity of this method, two experiments on torch height change and upward slope welding were carried out, and the result suggests that the proposed fuzzy controller had a fast dynamic response and good stability under every condition.