An ANFIS Based Approach for Predicting the Ultimate Bearing Capacity of Single Piles

The ultimate bearing capacity of single piles is an important design parameter for a single pile. Even though many methods have been proposed, they are based on field tests or mechanical analysis and there has not been an intelligence based approach for predicting the ultimate bearing capacity. This paper proposed, by applying the principles of the Adaptive Neuro-Fuzzy Inference System (ANFIS), an ANFIS based approach for predicting the ultimate bearing capacity of single piles. 42 examples of precast reinforced concrete piles were accumulated from an area in Shanghai, China. 37 examples of them were used to build a new prediction model for the ultimate bearing capacity of single piles. The model was then used to predict the ultimate bearing capacity of the remaining 5 examples. A comparison was conducted between the predicted results and the practical measurements. The comparison results show that the approach has the advantages of high speed of convergence, good capability of fitting, high accuracy of prediction and reproduction of the training process. The results also indicate that ANFIS based approach is an excellent approach for predicting the ultimate bearing capacity of single piles.