Soil Compaction Characteristics Modeling using Adaptive Neuro-fuzzy Inference System

Soil should be compacted to a desired density and water content in the construction of geotechnical structures. In other projects such as earth dams and compacted soil liners for containing contaminated solid and liquid wastes, the soil should be compacted for the density as well as the permeability requirements. This paper presents an adaptive neuro-fuzzy inference system (ANFIS) model to predict optimum moisture content (OMC) in different soil. To generate this model, a database consisting of 55 compaction test results was prepared and several ANFIS models were constructed to obtain the best one. Some statistical criteria consisting of root mean square error and coefficient of determination were used to check the model accuracy. In the ANFIS model, fineness modulus (FM), uniformity coefficient (U) and plastic limit (PL) were considered as model inputs. In ANFIS modelling, 44 datasets were considered for training purpose and 11 datasets were set for testing the model. The results indicate that the proposed ANFIS model is able to predict soil compaction characteristics with high degree of accuracy.

Soil compaction, optimum moisture content, adaptive neuro-fuzzy inference system