Neural network application in forecasting maximum wall deflection in homogenous clay

Abstract

Highlights Neural Networks was used to forecast maximum deflection of braced excavation in homogeneous clay and its position. A sensitivity analysis was accomplished to examine the relative significance of the parameters that influence the models. The results confirm that the developed ANN model is able to predict maximum deflection and its position reliably. Design charts were developed based on the ANN model.

Keywords

• Maximum wall deflection
• Braced excavation
• Homogeneous clay
• Neural network
• Forecasting