Prediction of dynamic soil properties based on subsurface information obtained from down-hole tests via intelligent methods

Soil dynamic properties are very influential in soil behavior. Subsurface information is available through various field experiments such as down-hole test. This study attempt’s to establish a relationship between shear wave velocity and soil parameters via intelligent methods such as support vector machine (SVM) and gene expression programming (GEP), after measuring the shear wave velocity by Down Hole experiments. Result showed that model A (III) with input parameters of VP, VS and γ led to more accurate results and using shear wave velocity (VS) caused an increment in model efficiency. In addition, the results showed that the support vector machine and gene expression planning have a high accuracy in predicting the soil Young's modulus and in the lack of information on soil properties, these parameters can be used to predict the Young's modulus of soil.

کلیدواژه ها:

soil properties ، young’s module ، shear wave velocity ، gene expression programming ، support vector machine ، intelligence methods