Prediction of Ultimate Bearing Capacity of Circular and Square Footings by Neural Network
Prediction of Ultimate Bearing Capacity of Circular and Square Footings by Neural Network
Mahdi Mollabagher Makhmalbaf1 Mohammad Azimipour2 Mahmoud Nikkhah3
1) MSc. Geotechnical Engineering, Islamic Azad University, Semnan Branch, Iran
2) MSc. Water & Wastewater Engineering, Shahid Beheshty University, Iran
3) Assistant Professor, Department of Civil Engineering, Islamic Azad University, Semnan Branch, Iran
IOI Code
:
XYBH-AZEFA
Publication :
International Conference on Civil Engineering, Architecture and Cityscape(iccacs.com)
Abstract :
For ultimate bearing capacity calculation, various Analytical – Experimental methods were proposed by Terzaghi, Mayerhof, Hansen and Vesic and the most comprehensive one in case of no axial load was proposed by Mayerhof. In this article Artificial Neural Network (ANN) is used to predict bearing capacity of circular and square shallow foundations and for this purpose Multi-layer Perceptron (MLP) is utilized. The data used as the inputs and output of network models were parameters in Mayerhof equation and the bearing capacity calculated from this equation, respectively. Finding the best architecture of model is carried out through making different conditions such as numbers of hidden neurons and activation functions. The result of this work shows high capabilities of this kind of neural network for bearing capacity prediction.
Keywords : bearing capacity mayerhof artificial neural network mlp