In this study, a method based on genetically optimized neural network system (GONNS) is introduced for selecting the optimized FSSW parameters. In this method for a known FSSW setup, type of material, and plate thickness, an artificial neural network (ANN) is designed with the process parameters as inputs and the weld tensile strength, maximum plunging load, and the process duration as outputs. Experiments are performed in order to train the selected ANN. Afterwards, an optimization algorithm is introduced based on the genetic algorithm heuristic search bundled with the trained ANN that aids the algorithm in predicting the value of the target functions. The target functions of the optimization problem are considered as normalized and weighted equations of the weld tensile strength, maximum plunging load, and the process duration.

Eventually, the minimization of the target function yields the optimum FSSW parameters which are verified by additional experiments. Results affirm that the analytically obtained optimum of the FSSW parameters is valid, and using the optimum parameters, the higher weld strength, lower plunging load, and shorter process duration are obtained.