Evaluation of the Compressibility Parameters of Soils Using Soft Computing Methods

Abstract

The compressibility parameters such as the compression index and the recompression index are necessary in the settlement calculation for fine-grained soils that is essential to geotechnical designs. However, determination of the compressibility parameters from odometer tests takes a relatively long time and leads to a very demanding experimental working program in the laboratory. Geotechnical engineering literature involves many studies based on multiple regression analysis (MLR). This study was aimed at predicting the compressibility parameters by soft computing methods such as artificial neural networks (ANN) and the quasi-Newton algorithm developed with the differential evolution method (QN-DE). The selected variables for each method are the index parameters of natural finegrained soils such as natural water content and initial void ratio. The results obtained from MLR, ANN, and QN-DE models were compared with each other at the end of the study.