Assessment of Geotechnical properties of fine grained soil using artificial intelligence techniques

Abstract

The stream of geotechnical engineering, which deals with understanding soil mechanics and newlinecharacteristics of soil, is called soil engineering. Soil engineering helps to analyze the newlinecomposition and formation process of soil. Each soil has different physical and chemical newlineproperties because of its formation process. In geotechnical engineering, the permeability, newlinestrength, and consistency limits are the physical properties of soils. In nature, many soils are newlinesticky, and many are not. The soil has sticky behavior, known as cohesive soil, and non-sticky newlinesoil is known as cohesionless soil. Each soil has its physical properties to study its behavior. newlineThe laboratory procedures are performed to obtain the physical properties of soil, which are newlinetime-consuming and arduous tasks. The permeability and strength tests require the optimum newlinemoisture content (OMC), determined experimentally from the standard proctor test. These newlineprocedures reveal the requirement to introduce new methods for evaluating soil properties. newlineSeveral investigators have introduced different methods for assessing soil properties to newlineescape from laboratory procedures and save time. The investigators have successfully newlineemployed soft computing and statistical methods to assess the soil properties, such as dry newlinedensity, California bearing ratio (CBR), moisture content, unconfined compressive strength newline(UCS), etc. The study of the reported work demonstrates that the investigators have employed newlineseveral soft computing techniques, i.e., random forest, ensemble tree, multivariate adaptive newlineregression splines, gaussian process regression, artificial neural network, decision tree, newlinemultivariable expression programming, support vector machine, etc., to assess the soil newlineproperties using a different number of databases. The published research reveals that soft newlinecomputing techniques are data-driven, and the accuracy and performance of soft computing newlinetechniques are affected by the quantity and quality of the database. newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced