كاربرد شبكه عصبي مصنوعي در مهندسي ژيوتكنيك مدل سازي تحليل و طراحي

The Application of Neural Networks in Geotechnical Engineering: Modeling, Analysis, and Design

Prediction of soil engineering behavior, based on previous experiments and observations, plays an important role in geotechnical engineering. So, to cover the subject, many researchers have focused on the mathematical and statistical models during the last decades. In this paper, multi-layer perceptron (MLP), a well-known method in Artificial Neural Networks, is used for modeling stress- strain behavior of silty sands, slop stability analysis of river banks and design of concrete liner of water -tunnels. In addition, MLP design and factor affecting its performance have been discussed briefly. Since all different kinds of problems have been covered (Modeling, Analysis and Design), authors see immense potential for the application of well-trained neural networks in geotechnical engineering.