Browsing by Author "Swamy, S.V."

Now showing 1 - 4 of 4

A machine learning framework for predicting elastic properties of sedimentary rocks from ball mill grinding characteristics data
(CRC Press/Balkema, 2024) Swamy, S.V.; Harish, P.; Kunar, B.M.; Chandar, K.R.
Elastic properties of rocks like Youngâ€™s modulus and compressional P-wave velocity are vital for understanding their stress-strain response in mining and rock engineering applications. Traditional methods for determining these properties involve labor-intensive, expensive and time-consuming. To address these challenges, this study proposes a novel predictive method. It utilizes a multi-layer perceptron feed forward neural network (MLP-FFNN) trained on grinding characteristics of ball mill to predict Youngâ€™s modulus and compressional Pwave velocity in sedimentary rocks. Laboratory experiments on limestone and dolomite samples generated extensive data, enabling development of prediction models using the proposed MLPFFNN. The developed models demonstrate high predictive accuracy (R values: 0.952 for E, 0.987 for Vp) in training and good generalization (0.866 for E, 0.9707 for Vp) in testing, along with low Root Mean Squared Error (RMSE) values. These findings underscore the efficacy of neural network models in predicting E and Vp from grinding characteristics of ball mill. Â© 2024 The Author(s).
Effect of longwall workings on the stability of overburden dumps
(CRC Press/Balkema, 2024) Harish, P.; Swamy, S.V.; Chandar, K.R.
Extraction of coal is done in both opencast and underground methods. Opencast mining generates a huge amount of overburden while excavating the coal. Managing the overburden material, deposited as dumps at considerable heights to minimize ground coverage, is crucial in opencast mines, but it poses risks such as potential failures. Such failures can stop the mining activities, endanger personnel safety, and damage equipment. At times, limitations in space, the placement of overburden dumps over underground excavations, posing stability challenges due to pre-existing stresses from activities below the surface. This paper explores the stability prediction of overburden dumps above longwall workings, using Rocscience RS2 v19.2, a two-dimensional finite element analysis software. Strength Reduction Technique determines the factor of safety (FOS), revealing that the presence of underground longwall excavation induces a vertical deformation of 56.4mm for the critical strength reduction factor of 1.12, emphasizing the impact on overburden dump stability. Â© 2024 The Author(s).
Predicting Rock Properties of Limestone Using Operating Parameters of Ball Mill
(Springer Nature, 2025) Swamy, S.V.; Kunar, B.M.; Chandar, K.R.
Rock properties are important for mining, geotechnical engineering, and other engineering projects. Accurate determination of these properties relies on high-quality samples, but challenges like sample availability, preparation of sample, cost, and time constraints have led to an increasing reliance on computational methods for prediction. Prior investigations predominantly relied on laboratory-based tests and indirect methodologies to predict properties of rocks. In contrast, this study introduces an innovative technique for predicting rock properties, specifically the P-wave velocity (Vp) and uniaxial compressive strength (UCS) by harnessing ball mill operational parameters throughout the grinding procedure an unconventional yet indirect approach. A multivariate regression model is established to connect operating parameters with the strength properties of limestone samples. The determination coefficients (R2) for Vp and UCS prediction models are 0.892 and 0.868, respectively. Moreover, an Analysis of Variance (ANOVA) is performed to ascertain the influence of significant parameters on the target variables. The accuracy and reliability of the prediction models are further validated through scatter plots and residual variations for both Vp and UCS models. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
Prediction of uniaxial compressive strength of limestone from ball mill grinding characteristics using supervised machine learning techniques
(Nature Research, 2025) Swamy, S.V.; Kunar, B.M.; Chandar, K.R.; Alwetaishi, M.; Shashikumar, S.; Reddy, S.
Uniaxial Compressive Strength (UCS) is a fundamental parameter in rock engineering, governing the stability of foundations, slopes, and underground structures. Traditional UCS determination relies on laboratory tests, but these face challenges such as high-quality core sampling, sample preparation difficulties, high costs, and time constraints. These limitations have driven the adoption of indirect approaches for UCS prediction. This study introduces a novel indirect method for predicting uniaxial compressive strength, harnessing the grinding characteristics of a ball mill as predictive variables through supervised machine learning techniques. The correlation between grinding characteristics and UCS was examined to determine whether a linear relationship exists between them. A hybrid support vector machine-recursive feature elimination (SVM-RFE) algorithm is applied to identify the critical grinding parameters influencing UCS. Four supervised machine learning models viz., Multiple Linear Regression (MLR), k-Nearest Neighbor Regression (k-NNR), Support Vector Regression (SVR), and Random Forest Regression (RFR) were developed for UCS prediction, with hyperparameter optimization performed using RandomisedSearchCV technique. The Random Forest model outperformed others as the best prediction model, achieving a coefficient of determination (R²) of 0.95, followed by SVR (R² = 0.87), k-NNR (R² = 0.82), and MLR (R² = 0.758). Model robustness was further assessed using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Variance Accounted For (VAF). Internal validation by means of K-fold cross validation and external validation with independent datasets confirmed generalization capability, showing an average prediction error of ± 10%. The findings demonstrate that combining grinding characteristics with machine learning offers an accurate, cost-effective alternative to conventional UCS testing, with significant practical applications in rock engineering. © The Author(s) 2025.