Prediction of Toe Stability in Rubble Mound Breakwaters Using Artificial Neural Networks

Rubble mound breakwaters with concrete and stone armor units are essential in coastal engineering, protecting against wave action and sediment transport. Stability of toe in these structures is critical, supporting the armor layer and preventing scour, thus influencing the breakwater's performance and longevity. This study advances toe stability prediction using an Artificial Neural Network (ANN) approach. Experimental data from laboratory tests on breakwaters with concrete, stone, and composite toes, conducted mainly in deep water conditions, were used to train and validate the ANN model. Featuring a hidden layer and an advanced optimization algorithm, the model effectively captures complex, non-linear relationships. The ANN's performance was compared to empirical methods using metrics such as the correlation coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), mean squared error (MSE), correlation coefficient (CC), and normalized bias (nBias). Results demonstrated the ANN's superior accuracy, with higher R۲ and lower RMSE, MAE, MSE, and nBias values, highlighting the significant potential of neural networks for enhancing toe stability analysis in breakwater design.

کلیدواژه ها:

Rubble mound breakwater ، toe stability ، concrete toe ، artificial neural network