Application of the A.N.Ns to Estimate the Reflection and Transmission Coefficients of Pontoon Floating Breakwaters

One of the main problems in valuation of the operation of floating breakwaters is estimation of thetransmission and reflection coefficients and because of absence the universal theoretical principles in order to obtain these coefficients in any region, investigations have been focused on physical models. In this research program, A.N.Ns (Artificial Neural Networks) has been used as a computational tool in order to estimate the transmission and reflection coefficients of the P.F.Bs (pontoon floating breakwaters). The mathematical A.N.N model used in this study, is a MLP (Multi Layer Perception) kind and BP (Back Propagation) algorithm has been used for learning the network. The data from physical model studies of P.F.B are partly used for learning the network and the remained are used to verify the learned network. The different parameters of neural networksincluding the learning rule, transfer function, number of hidden layers and number of nodes in each layer have been investigated and finally, the optimum network, have been determined.