Optimization of parameters of pulse current gas tungsten arc welding using non conventional techniques

Abstract

This paper explains the heuristic search and evolutionary methods with new approach using experimental design techniques i.e Central composite design. The experimental data collected using conventional experimentation is used in application of genetic algorithm (GA) and particle swarm optimization (PSO) for optimization of process parameters. The data available from the experimentation is being used to perform genetic algorithm and particle swarm optimization and is being compared with the predicted values using design of experiment i.e from regression model. The developed regression model of the objective function i.e properties of weld joint also act as the objective function for optimization through genetic algorithm and particle swarm optimization. Modelling and optimization of process variables to find the quality weld output using pulse current GTAW has been tried in the present work. The optimized values obtained from these techniques were compared with experimental results and presented.