Application of Neural networks in prediction of beams natural modes

The objective of this study is to discuss the sufficiency of NN as a securer, quicker and more robust method to be used in prediction of natural mode beams. For this purpose, one ANN models was proposed to estimate third natural mode beam with L cross section. In the ANN models, a back-propagation (BP) algorithm was employed using a scaled conjugate gradient (SCG). ANN model was developed, trained and tested in a based MATLAB program. A training set of 184 and a validation set of 14 beams were produced from dynamic response of different beams. Finite Element Analysis (FEA) in based ANSYS program was used to generate training and testing set of ANN models. It was demonstrated that the neural network based approach is highly successful to determine third natural mode.