Prediction of Hydrodynamic Coefficients of Stratified Porous Structure Using Artificial Neural Network (ANN)

Abstract

The breakwater is designed to offer tranquility in the harbor to protect the offshore facilities and also to prevent coastal erosion. The use of soft computing approaches in coastal engineering helps to solve the nonlinear problems and predicts the hydrodynamic performance of the device. In the present study, the artificial neural networks (ANNs) with different topologies are considered to predict the hydrodynamic coefficients for the wave interaction with the stratified porous breakwater. The experimental study is performed to determine the reflection and transmission coefficient for the horizontally stratified porous structure with three layers of different porosity and width of the structure. The hydrodynamic performance is analyzed by considering the feed-forward back-propagation neural network, and the results are compared for different numbers of hidden nodes. Further, the root mean square error (RMSE) and coefficient of correlation (CC) are considered to assess the ability of ANN topologies to predict the transmission coefficient. The numerical results obtained using ANN are noted to fall within the range that represents the network’s ability to predict accurate results. The study performed will provide an insight in the design and analysis of the stratified porous breakwater in the nearshore regions. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.