Numerical, ANN, and MLR Modelling for Stability Analysis of Coal Mine Overburden Dump
摘要
Overburden (OB) dump slope handling has been critical to the effectiveness of the mining operation. Several cases of dump failures are reported due to the inappropriate geometrical configuration of the dump. The current study examined the stability of the external dump slope of two coal mines in the Indian state of Chhattisgarh under various geometrical configurations. To assess the various geotechnical properties of the OB material from both mines, laboratory studies were undertaken. Numerical modelling was used to examine the stability of coal mine exterior dumps in various geometrical configurations using FLAC/ Slope software. The exterior dump height was varied from 30, 40, 50, and 60 m, and the dump slope from 30° at 2° intervals. Furthermore, the optimal dump slope of external coal mines has been recommended based on the required Factor of Safety (FoS) Value. An Artificial Neural Network (ANN) and Multiple Linear Regression (MLR) models were also established for the prediction of FoS values based on the different input parameters i.e., dump slope, dump height, unit weight, cohesion, and friction angle. The resulting coefficient of determinations (R2 > 0.90) values of ANN and MLR models suggest that both the models have a high degree of statistical reliability.