Welding strength prediction in nuts to sheets joints: machine learning and ANFIS comparative analysis

This study uses machine learning algorithms and the Adaptive Neuro-Fuzzy Inference System (ANFIS) to predict welding strength in DD13 sheet metal joints with AISI 1010 nuts. The objective is to optimize industrial welding processes and improve quality control. The study investigates weld current, time, and hold time as critical input variables for joint integrity. The performance of different ML algorithms, including linear regression, random forest regression, ridge regression, Bayesian regression, K-Nearest Neighbors regression, decision tree regression, and ANFIS, are evaluated. Training and testing data consist of welding parameters and corresponding strength measurements. Performance metrics such as R2 score, mean absolute error (MAE), mean squared error (MSE), and root mean square error (RMSE) are used to assess the predictive capabilities. Random forest regression is the most efficient algorithm, with a high R2 score of 0.992 and minimal errors. ANFIS also exhibits comparable performance, highlighting its efficacy in this context. These findings can be useful for optimizing welding parameters in industrial settings, potentially leading to improved quality control and weld strength, particularly in automotive applications. Using ML and ANFIS, industries can make informed decisions to optimize welding processes and ensure joint integrity, ultimately meeting the rigorous demands of demanding applications.