Prediction Of Ultimate Pile Bearing Capacity Using Artificial Neural Networks

The behavior similarity between pile and cone penetration test, has guided the researchers to use the CPT results in prediction of piles bearing capacity. In this article, artificial neural networks have been used to study about relation between cone tip and sleeve friction strength obtained from cone penetration test and ultimate pile bearing capacity, obtained from static pile load test. Artificial neural networks "ANNS" is a form of artificial intelligence. ANN tries to dramatize the human nerves & brain system biology structure in self-structure. In this article, the results of relevant studies about the prediction of ultimate pile bearing capacity in especial area of soils are presented by use of back - propagation neural networks. The results of static pile load test up to failure on 63 piles with closed end and CPT test results in the place of construction these piles use in order to confirm and creation ANN model. Comparing predicted ultimate pile bearing capacity by utilizing ANN with the numbers which are predicted by 5 common traditional methods shows that using of artificial neural networks is a suitable method for predicted ultimate pile bearing capacity in identified area of texture of soil moreover it has a better operation in comparison with common traditional methods.