DefinePK

Abstract

In this paper, Gene Expression Programming (GEP) models are developed to select the appropriate ground improvement technique. Ground improvement is usually used to increase the soil bearing capacity, reduce potential settlement, and mitigate liquefaction. The data used to build the GEP models was collected from 83 ground improvement projects in the UAE. Data collected from each project includes the following parameter: fine content (%), groundwater level (m), depth of improvement (m), distance to close by structures (m), and the used ground improvement method. This paper investigates five ground improvement techniques that are dynamic compaction (DC), dynamic replacement (DR), vibro compaction (VC), rapid impact compaction (RIC), and stone columns (SC). One GEP model is developed for each technique, the user will be able to input the above-mentioned parameter in each model and select the technique with higher accuracy. The developed GEP models have R2 values, for the training dataset, ranging from 0.72 to 0.95. The results showed that GEP could select the appropriate ground improvement technique with acceptable accuracy.