Estimation of the rock deformation modulus and RMR based on data mining techniques

• Author(s): Martins, Francisco F. ; Miranda, Tiago F. S.
• Date: 2012
• Persistent ID: http://hdl.handle.net/1822/20085
• Origin: RepositóriUM - Universidade do Minho
• Subject(s): Deformation modulus; RMR; Data mining; Machine learning
• Document Type: Article
• Language: English

Description

In this work Data Mining tools are used to develop new and innovative models for the estimation of the rock deformation modulus and the Rock Mass Rating (RMR). A database published by Chun et al. (2008) was used to develop these models. The parameters of the database were the depth, the weightings of the RMR system related to the uniaxial compressive strength (UCS), the rock quality designation (RQD), the joint spacing (JS), the joint condition (JC), the groundwater condition (GWC) and the discontinuity orientation adjustment (DOA), the RMR and the deformation modulus. As a modelling tool the R program environment was used to apply these advanced techniques. Several algorithms were tested and analysed using different sets of input parameters. It was possible to develop new models to predict the rock deformation modulus and the RMR with improved accuracy and, additionally, allowed to have an insight of the importance of the different input parameters.