1. Stability analysis and artificial neural network as result validation in crusher, MSE wall, and stockpile interaction: A study case.
- Author
-
Vergiagara, Vega, Chusna, Ifa Aulia, Saptono, Singgih, and Haq, Shofa Rijalul
- Subjects
- *
ARTIFICIAL neural networks , *REINFORCED soils , *BACK propagation , *SLOPES (Soil mechanics) , *SAFETY factor in engineering - Abstract
The crusher facility is recognized as being a principal preparation facility for processing and refining minerals. Installing the crusher requires a steep slope to convenience the feeding process. Currently, Mechanically Stabilized Earth (MSE) Wall has been decided to reinforce a steep slope on the crusher facility to increase stability. The crusher vibration and stockpile load on the crusher area widely account for the crusher facility's longs term stability. This paper discusses how to define the interaction of the crusher, MSE Wall, and stockpile that affects the crusher stability. GEO5 software has provided the Limit Equilibrium Method (LEM) in two-dimensional analysis using Mohr-Coulomb Failure Criteria and Morgenstern-Price Method. Back Propagation Artificial Neural Network (ANN) is used to validate the computational analysis from GEO5. The sum of variations models in this analysis was 25 slopes. Based on the variations, the data input for ANN included 20 models, while the five models are used to validate the ANN prediction result. The entire validation model meets the criteria for the value of R2 > 0.6. Still the same as the training model where only parameter Center of Rotation: y Axis Values has an SI of RMSE above 10%, which is 13.6%, which means that the neural network model is used in this study is still not able to estimate the shape of the Center of Rotation, however this Neural Network model can excellently estimate the value of the Safety Factor because based on the R2 value is 0.96 and the SI of RMSE is 0.59%. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF