AN EFFICIENT CALIBRATION PROCESS FOR THE PREDICTION OF ROCK STRENGTH THROUGH MACHINE LEARNING ALGORITHMS

Authors : Şaziye Özge DİNÇ GÖĞÜŞ

Pages : 311-326

Doi:10.59313/jsr-a.1239780

View : 28 | Download : 10

Publication Date : 2023-03-29

Article Type : Research Paper

Abstract :Numerical models based on the discrete element method insert ignore into journalissuearticles values(DEM); have been widely used to predict the mechanical behaviors of rocks in rock engineering applications. Nevertheless, calibration of the model parameters is done by running numerous simulations and this time-consuming simulation process precludes the numerical platforms to be used as a practical tool in such applications. This study aims to accelerate the calibration process of the micro-parameters of three-dimensional insert ignore into journalissuearticles values(3D); numerical models built based on DEM and facilitate the generation of an efficient database by using machine learning algorithms in the prediction of rock strength. Namely, these algorithms are linear regression insert ignore into journalissuearticles values(LR);, decision tree insert ignore into journalissuearticles values(DT); regression, and random forest insert ignore into journalissuearticles values(RF); regression. The appropriate methodology for predicting the uniaxial compressive strengths insert ignore into journalissuearticles values(UCS); of certain rock types was investigated using a dataset consisting of micro-parameters of 87 DEM-based rock models, generated through an open-source code, Yade. The performance of such methods was evaluated by using metrics including R-squared score insert ignore into journalissuearticles values(R2);, mean squared error insert ignore into journalissuearticles values(MSE);, root mean squared error insert ignore into journalissuearticles values(RMSE);, and mean absolute error insert ignore into journalissuearticles values(MAE);, and then their statistical discrepancies were analyzed. The most accurate prediction of UCS was obtained in the LR method and the lowest percentage of performance was derived from the RF algorithms. LR method provides the results efficiently during calibration of the micro-parameters of a DEM-based rock model.
Keywords : Rock strength, Discrete element method, Numerical model calibration, Machine learning

ORIGINAL ARTICLE URL