Modeling load-settlement behavior of driven piles embedded in cohesive soils using artificial neural networks

Abstract

An accurate prediction of pile load-settlement behavior under axial load is necessary for design. This paper presents the development of a new model to predict the load-settlement behavior of pile foundations driven into cohesive soils and subjected to axial loads. Artificial neural networks (ANNs) have been utilized for this purpose. The data used for development of the ANN model is collected from the literature and comprise a series of in-situ driven piles load tests as well as cone penetration test (CPT) results. The data are divided into two subsets: Training set for model calibration and independent validation set for verification the performance of the ANN model in the real world. Sequential neural network is used for modeling. Predictions from the ANN model are compared with the results of experimental data and statistical measures are used to verify the performance of the model. The results indicate that the ANN model performs very well and able to predict the pile load-settlement relationship accurately.