Modeling of the MIG Welding Process Using Interval Type-2 Fuzzy Logic Systems

• DOI: https://doi.org/10.22533/at.ed.3175825011011

• Palavras-chave: MIG welding, type-2 fuzzy logic, uncertainty

• Keywords: process modeling, intelligent control

• Abstract: Abstract This study presents a model based on Interval Type-2 Fuzzy Logic Systems (IT2 FLS) applied to the Metal Inert Gas (MIG) welding process, aimed at representing and predicting weld bead behavior under uncertain conditions. The model considers six input variables: gas flow rate, work angle, wire feed speed, arc voltage (Trim), travel speed (IPM), and welding technique (push or pull), using experimental data obtained from an automated welding cell. Each variable was characterized by five membership functions (very low, low, medium, high, and very high), allowing the definition of the Footprint of Uncertainty (FOU) associated with experimental dispersion. Statistical validation through Analysis of Variance (ANOVA) showed a value of p=0.0104<0.05 , confirming the statistical significance of the model, while the coefficient of determination R2=0.8640 indicated that the model explained 86.4% of the total variability in the experimental data. These results demonstrate that the IT2 FLS is a robust and accurate tool for modeling welding processes with nonlinear and uncertain behavior, providing a solid foundation for the development of intelligent systems for control and optimization of the MIG process.