Modelling Neural Networks for prefiguration of the tensile strength of Friction Stir Welded Pure Copper joints

摘要

Artificial Neural Network (ANN) possesses a remarkable ability to extract connotation from different set of data structures. It is inspired from the mimicking of the working of biological nervous system. ANN learning abilities are more like us because they learn by examples. In this research paper prefiguration of the tensile strength of Friction Stir Welded pure copper alloys is performed. The Quasi- Newton algorithm method is used for training the neural networks. The results showed that the traverse speed is most important variable which contributes 101.3% to the output i.e. tensile strength. The accuracy of 95.71% is obtained between the actual tensile strength and predicted tensile strength......