Publication

The paper titled “Post Weld Heat Treatment Optimization of Dissimilar Friction Stir Welded AA2024-T3 and AA7075-T651 Using Machine Learning and Metaheuristics” was published in the journal Materials, Volume 16, Issue 5, and is indexed in the Scopus database (Quartile 2).

This research focuses on optimizing post weld heat treatment (PWHT), a process used to enhance the mechanical properties of welded materials. While past studies have examined PWHT effects using experimental methods, this work introduces a novel approach by integrating machine learning (ML) and metaheuristics for both modeling and optimization, advancing intelligent manufacturing applications.

The study employs ML techniques—support vector regression (SVR), K-nearest neighbors (KNN), decision tree (DT), and random forest (RF)—to model the relationship between PWHT parameters and key mechanical outcomes: ultimate tensile strength (UTS) and elongation percentage (EL). Results show SVR outperforms other methods and, when paired with metaheuristics such as differential evolution (DE), particle swarm optimization (PSO), and genetic algorithms (GA), SVR-PSO achieves the fastest convergence. The research ultimately suggests both single-objective and Pareto-optimal solutions for PWHT parameter settings.