Prediction of Wave Reflection for Quarter Circle Breakwaters Using Soft Computing Techniques

The modified form of the semi-circular breakwater is called Quarter-Circle Breakwater (QBW). It consists of a quarter-circular surface facing incident waves, a horizontal bottom, a rear wall, and is built on a rubble mound foundation. In general, QCB may be constructed as emerged, with and without perforations that may be on one side or either side based on the coastal designer. These perforations dissipate the energy due to the formation of eddies and turbulence created inside the hollow chamber. In the present study, experimental data obtained from Binumol, 2017 are fed as input to both the models. This data is used to predict the reflection coefficient of QBW by adopting the ANN system approach. The reliability of the Artificial Neural Network (ANN) approach is done with statistical parameters, namely Model Performance Analysis (MPA) viz., Correlation Coefficient (CC), Root Mean Square Error (RMSE), Nash-Sutcliffe Efficiency (NSE), and Scatter Index (SI). The results of the MPA indicate that the ANN is suited for predicting the reflection coefficient of QBW.