Application of genetic algorithm and particle swarm optimization in the optimization of mild steel weld parameters

Optimizing the welding process is critical in enhancing the weld quality, its efficiency, and cost-effectiveness. The process of finding the optimal weld output, is inherently complex considering interconnected factors. This study explores the application of two metaheuristic techniques; Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), in optimizing the weld strength of a Gas Tungsten Arc Welding (GTAW) mild steel plate. Results showed that GA achieved optimal weld conditions with a tensile strength of 459.53 N/mm², penetration depth of 2.71 mm, and bead width of 3.19 mm. However, PSO with a tensile strength of 491.23 N/mm², penetration depth of 2.23 mm, and bead width of 3.51 mm, outperformed the GA. The study reveals PSO’s superior performance under the given welding parameters, with the PSO achieving faster convergence and requiring less computational effort compared to the GA. Further study on the effect of the weld parameters using the 3D surface plots, reveals a stronger dependence of the tensile strength on the weld current and root gap, with a noticeable curvature suggesting optimal combinations at moderate current and root gap values.