Prediction of Surface Deformation for Open Cast Mine Based on Pit Wall Prisms Monitoring Dataset
Abstract
This paper explores the application of Artificial Intelligence (AI) techniques for predicting pit wall deformation in open-cast mining operations. Four AI models, including the Patient Rule Induction Method (PRIM), Radial Basis Function Neural Network (RBFNN), Back Propagation Neural Network (BPNN), and Group Method of Data Handling (GMDH), were developed and evaluated to estimate pit wall deformation for three different monitoring locations designated as Prisms 1, 2 and 3. The AI models were statistically evaluated using dimensioned error indicators such as Mean Square Error (MSE), Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The coefficient of determination (R2) was also used for the model’s performance evaluation. The study concluded that for the Prism 1 monitoring point, the BPNN was the most suitable for predicting the pit wall deformation. However, for Prisms 2 and 3, the RBFNN demonstrated superior performance, with minimal errors and high R2 scores, making it a suitable choice for deformation prediction. GMDH exhibited fair results, while PRIM produced significant prediction errors, rendering it less suitable for pit wall deformation estimation. In general, the study findings suggest that AI techniques can significantly enhance and automate the deformation prediction process in open-cast mining, offering opportunities for improved safety and operational efficiency.
Downloads
Published
Issue
Section
License
Copyright (c) 2024 Yakubu Issaka, Yao Yevenyo Ziggah, Frank Osei
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright © 2021 University of Mines and Technology (UMaT), Tarkwa. Ghana