Comparison between genetic algorithms and response surface methodology in GMAW welding optimization

The response surface methodology (RSM) is a traditional technique for experimental process optimization. Recently, a new approach to this problem has been tried with the genetic algorithm (GA), which is most known in the numerical field. The present paper compares these two techniques in the optimization of a GMAW welding process application. The situation was to choose the best values of three control variables (reference voltage, wire feed rate and welding speed) based on four quality responses (deposition efficiency, bead width, depth of penetration and reinforcement), inside a previous delimited experimental region. For the RSM, an experimental design was chosen and tests were performed in order to generate the proper models. In the GA case, the search for the optimal was carried out step by step, with the GA predicting the next experiment based on the previous, and without the knowledge of the modeling equations between the inputs and outputs of the GMAW process. Results indicate that both methods are capable of locating optimum conditions, with a relatively small number of experiments.