Slope Stability Prediction of Homogenous Earth Dam Caused by Fluid Particles Seeps by Using Artificial Neural Networks

Abstract

Artificial neural networks (ANN), neural networks or artificial intelligence, which can model complex functions. They are useful in predicting the output of two or more independent variables. Further, forecasting the stability of slopes of earth embankment is a very exciting task for civil engineers due to the geology of the region, shear strength, and groundwater of earth embankment in accessing slope constancy. In this paper, a prediction process has been developed for predicting the safety factor (FOS) of slopes by using the ANN. 243 circumstances with dissimilar arithmetical, soil factors of homogenous earth dam and different fluid heights were analysed using multilayer method. Out of these, 70% of cases were used for training the model and 30% for testing the model. The results showed that if we used 70% of testing and 30% for training with one hidden layer is the best choice with R=0.901, in contrast of using all cases with testing, which goes far away from unity.