Optimization of resistance spot welding parameters using differential evolution algorithm and GRNN

Welding is a basic manufacturing process for making components or assemblies with good mechanical properties. Resistance spot welding (RSW) is used frequently as a successful joining method for a variety of work commonly in automotive and other manufacturing processes. Because of complicacy during the RSW and lots of interferential factors, especially short-time property of the process, it is very difficult to build a mathematical model that can predict the output accurately. This paper presents a novel technique based on general regression neural network to approximate the relationship between welding parameters (welding current, electrode force, welding time and metal sheet thickness) and the failure load that can withstand the joint. A model is formulated from the trained experimental data through general regression neural network. Differential Evolution Algorithm is then applied on to the model to obtain the optimum combination of welding parameters to offer better weld joint strength at low power consuption.