Application of ANFIS Adaptive System to Estimate the Potential Consolidation of Clay Soils

The consolidation phenomenon occurs in the construction of structures on saturated clay soils which in effect this phenomenon the soil particles is compressed. In sometimes an irreparable damage occurs in the development projects affected of the soil settlement. Hence the prediction of soil settlement is necessary. The compression index (Cc) is one of the coefficients which apply to calculations of soil settlements. This coefficient is determined by odoemeter test which is test is very expensive and much time is wasted. So that in past years an extensive function for prediction of Cc with physical properties of soil have been developed by various researchers. In this study, the important and valuable results of consolidation test were gathered at different locations in khuzestan province of Iran. Using the gathered data the compression index was determined. Then the results were evaluated using the simultaneous implementation technique of artificial neural network and fuzzy logic (ANFIS). In this research, also, the results of ANFIS system were compared to the results of empirical formula. The error of empirical functions to measured data in the best cases was predicted about 20 percent. However, the most of empirical relationships in the study area have not acceptable accuracy. Findings show that ANFIS adaptive system predicates compression index which it is a fundamental parameter for determining of clay soil settlement. The results of this system were satisfactory with void ratio and wet content variable in situ clay soils. In addition, this system with 15 percent RMSE and 12 percent error has a better result than past empirical function and artificial neural network.