Rock Mechanics
Kapoor Chand; Ved Kumar; Priyanshu Raj; Nikita Sharma; Amit Kumar Mankar; Radhakanta Koner
Abstract
Failure of tailings dams is a major issue in the mining industry as it critically impacts the environment and life. A major cause of the failure of tailings dams is the unplanned depositing of tailings and the increase in saturation due to rainfall events. This study using numerical modelling and artificial ...
Read More
Failure of tailings dams is a major issue in the mining industry as it critically impacts the environment and life. A major cause of the failure of tailings dams is the unplanned depositing of tailings and the increase in saturation due to rainfall events. This study using numerical modelling and artificial intelligence techniques (like MLR, SVR, DT, RF, and XGB) aims to predict the slope stability of tailings dams to avoid failure. The stability of tailings dams is analysed using the finite difference method (FDM), which computes the factor of safety (FoS) using the shear strength reduction (SSR) technique. This investigation mainly focuses on the geotechnical and geometric parameters of the tailings dam, such as density, cohesion, friction angle, saturation, embankment height, slope angle and haul road width. Results of numerical modelling have been used for developing ML models and predicting slope stability. The efficiency of ML models was analysed based on the R2 and root mean square error (RMSE), mean squared errors (MSE), and mean absolute error (MAE). The XGB algorithm proved to be the most effective as it gave the highest accuracy and lowest RMSE value compared to other ML models. AI tool was developed based on the ML model results for dam slope stability prediction. The developed AI tool will help understand the role of saturation and geometry parameters in embankment stability at the initial level of investigation.
Exploration
Moslem Jahantigh; Hamid Reza Ramazi
Abstract
Fuzzy c-means (FCM) is an unsupervised machine learning algorithm. This method assists in integrating airborne geophysics data and extracting automatic geological map. This paper tries to combine airborne geophysics data consisting of aeromagnetic, potassium, and thorium layers to classify the lithological ...
Read More
Fuzzy c-means (FCM) is an unsupervised machine learning algorithm. This method assists in integrating airborne geophysics data and extracting automatic geological map. This paper tries to combine airborne geophysics data consisting of aeromagnetic, potassium, and thorium layers to classify the lithological map of the Shahr-e-Babak area, a world-class porphyry area in the south of Iran. The resulting clusters with FCM show appropriate coincidence with the geological map of the study area. The clusters are adapted with high magnetic anomalies corresponding to the mafic volcanic rocks and the clusters with high radiometric signature associated with igneous rocks. The cluster is associated with low magnetic anomaly and low radioelements concentration representing sedimentary rocks. some clusters are associated with two or more lithological formations due to similar signatures of geophysics properties. The fuzzy score membership in all clusters is above 0.71 indicating a high correlation between geological signatures and multigeophysical data. This study shows geophysical signatures analyzed with the machine learning method can reveal geological units.
Exploitation
Marco Antonio Cotrina Teatino; Jairo Jhonatan Marquina Araujo; Eduardo Manuel Noriega Vidal; Jose Nestor Mamani Quispe; Johnny Henrry Ccatamayo Barrios; Joe Alexis Gonzalez Vasquez; Solio Marino Arango Retamozo
Abstract
The primary objective of this research was to apply machine learning techniques to predict the production of an open pit mine in Peru. Four advanced techniques were employed: Random Forest (RF), Extreme Gradient Boosting (XGBoost), K-Nearest Neighbors (KNN), and Bayesian Regression (RB). The methodology ...
Read More
The primary objective of this research was to apply machine learning techniques to predict the production of an open pit mine in Peru. Four advanced techniques were employed: Random Forest (RF), Extreme Gradient Boosting (XGBoost), K-Nearest Neighbors (KNN), and Bayesian Regression (RB). The methodology included the collection of 90 datasets over a three-month period, encompassing variables such as operational delays, operating hours, equipment utilization, the number of dump trucks used, and daily production. The data were allocated 70% for training and 30% for testing. The models were evaluated using metrics such as Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Variance Accounted For (VAF), and the Coefficient of Determination (R2). The results indicated that the Bayesian Regression model was the most effective in predicting production in the open pit mine. The RMSE, MAPE, VAF, and R2 for the models were 3686.60, 3581.82, 4576.61, and 3352.87; 12.65, 11.09, 15.31, and 11.90; 36.82, 40.72, 1.85, and 47.32; 0.37, 0.41, 0.41, and 0.47 for RF, XGBoost, KNN, and RB, respectively. This research highlights the efficacy of machine learning techniques in predicting mine production and recommends adjusting each model's parameters to further enhance outcomes, significantly contributing to strategic and operational management in the mining industry.
Environment
Podicheti Ravi Kiran; Ramchandar Karra
Abstract
Opencast coal mines play a crucial role in meeting the energy demands of a country. However, the operations will result in deterioration of ambient air quality, particularly due to particulate emissions. The dispersion of particulate matter will vary based on the mining parameters and local meteorological ...
Read More
Opencast coal mines play a crucial role in meeting the energy demands of a country. However, the operations will result in deterioration of ambient air quality, particularly due to particulate emissions. The dispersion of particulate matter will vary based on the mining parameters and local meteorological conditions. There is a need to establish a suitable model for predicting the concentration of particulate matter on a regional basis. Though a number of dispersion models exist for prediction of dust concentration due to opencast mining, machine learning offers several advantages over traditional modeling techniques in terms of data driven insights, non-linearity, flexibility, handling complex interactions, anomaly detection, etc. An attempt has been made to assess the dispersion of particulate matter using machine learning techniques by considering the mining and meteorological parameters. Historical data comprising of mine working parameters, meteorological conditions, and particulate matter pertaining to one of the operating opencast coal mines in southern India has been utilized for the study. The data has been analyzed using different machine learning techniques like bagging, random forest, and decision tree. The performance metrics of test data are compared for different models in order to find the best fit model among the three techniques. It is found that for PM10, many of the times bagging technique gave a better accuracy, and for PM2.5, decision tree technique gave a better accuracy. Integration of mine working parameters with meteorological conditions and historical data of particulate matter in developing the model using machine learning techniques has helped in making more accurate predictions.
Exploration
Naman Chandel; Sushindra Kumar Gupta; Anand Kumar Ravi
Abstract
Groundwater is an essential resource for human survival, but its quality is often degraded by the human activities such as improper disposal of waste. Leachate generated from landfill sites can contaminate groundwater, causing severe environmental and health problems. Machine learning techniques can ...
Read More
Groundwater is an essential resource for human survival, but its quality is often degraded by the human activities such as improper disposal of waste. Leachate generated from landfill sites can contaminate groundwater, causing severe environmental and health problems. Machine learning techniques can be used to predict groundwater quality and leachate characteristics to manage this issue efficiently. This study proposes a machine learning-based model for the prediction of groundwater quality and leachate characteristics using the effective water quality index (EWQI). The leachate dataset used in this study was obtained from a landfill site, and the groundwater quality dataset was collected from literature review. The mean values of TDS, Ca, Mg, NO3-, and PO4- exceeded the prescribed limit for drinking water purposes. The proposed model utilizes a machine learning architecture based on a convolutional neural network (CNN) to extract relevant features from the input data. The extracted features are then fed into a fully connected network to estimate the EWQI of the input samples. The model, trained and tested on leachate and groundwater quality datasets, achieves a high accuracy and computational efficiency, aiding in predicting groundwater quality and leachate characteristics for waste management.
Exploration
Ajay Kumar
Abstract
Land use (LU) classification based on remote sensing images is a challenging task that can be effectively addressed using a learning framework. However, accurately classifying pixels according to their land use poses a significant difficulty. Despite advancements in feature extraction techniques, the ...
Read More
Land use (LU) classification based on remote sensing images is a challenging task that can be effectively addressed using a learning framework. However, accurately classifying pixels according to their land use poses a significant difficulty. Despite advancements in feature extraction techniques, the effectiveness of learning algorithms can vary considerably. In this study conducted in Talcher, Odisha, India, the researchers proposed the use of Artificial Neural Networks (ANNs) to classify land use based on a dataset collected by the Sentinel-2 satellite. The study focused on the Talcher region, which was divided into five distinct land use classes: coal area, built-up area, barren area, vegetation area, and waterbody area. By applying ANNs to the mining region of Talcher, the researchers aimed to improve the accuracy of land use classification. The results obtained from the study demonstrated an overall accuracy of 79.4%. This research work highlights the importance of utilizing remote sensing images and a learning framework to address the challenges associated with pixel-based land use classification. By employing ANNs and leveraging the dataset from the Sentinel-2 satellite, the study offers valuable insights into effectively classifying different land use categories in the Talcher region of India. The findings contribute to the advancement of techniques for accurate land use analysis, with potential applications in various fields such as urban planning, environmental monitoring, and resource management.
Yahia ElSayed Khamis; Shady Galal El-Rammah; Adel M Salem
Abstract
The rate of penetration plays a key role in maximizing drilling efficiency, so it is essential for the drilling process optimization and management. Traditional mathematical models have been used with some success to predict the rate of penetration in drilling. Due to the high complexity and non-linear ...
Read More
The rate of penetration plays a key role in maximizing drilling efficiency, so it is essential for the drilling process optimization and management. Traditional mathematical models have been used with some success to predict the rate of penetration in drilling. Due to the high complexity and non-linear behavior of drilling parameters with the rate of penetration, these mathematical models cannot accurately and comprehensively predict the rate of penetration. Machine learning (ML) seems to be an attractive alternative to model this complicated physical process. This research paper introduces new data-driven models used to predict ROP using different parameters such as (depth, weight on bit (WOB), revolution per minute (RPM), Torque (T), standpipe pressure (SPP), flow in pump (pumping flow rate(Q), mud weight, hours on bit (HOB), revolutions on bit, bit diameter, total flow area (TFA), pore pressure, overburden pressure, and pit volume). Data-driven models are built using two different machine learning techniques, using 1771 raw real field data. The coding is built using the python programming language. The k-nearest neighbors (KNN) model predicting ROP for the training dataset show a correlation coefficient (R2) of 0.94. The multi-layer perceptron (MLP) model predicting ROP for the training dataset show a correlation coefficient (R2) of 0.98. We can conclude that MLP has a better accuracy, and removing outliers enhances model performance.
Blessing Olamide Taiwo; Gebretsadik Angesom; Yewuhalashet Fissha; Yemane Kide; Enming Li; Kiross Haile; Oluwaseun Augustine Oni
Abstract
Rock blast production rate (BPR) is one of the most crucial factors in the evaluation of mine project's performance. In order to improve the production of a limestone mine, the blast design parameters and image analysis results are used in this work to evaluate the BPR. Additionally, the effect of rock ...
Read More
Rock blast production rate (BPR) is one of the most crucial factors in the evaluation of mine project's performance. In order to improve the production of a limestone mine, the blast design parameters and image analysis results are used in this work to evaluate the BPR. Additionally, the effect of rock strength on BPR is determined using the blast result collected. In order to model BPR prediction using artificial neural networks (ANNs) and multivariate prediction techniques, a total of 219 datasets with 8 blasting influential parameters from limestone mine blasting in India are collected. To obtain a high-accuracy model, a new training process called the permutation important-based Bayesian (PI-BANN) training approach is proposed in this work. The developed models are validated with new 20 blast rounds, and evaluated with two model performance indices. The validation result shows that the two model results agree well with the BPR practical records. Additionally, compared to the MVR model, the proposed PI-BANN model in this work provides a more accurate result. Based on the controllable parameters, the two models can be used to predict BPR in a variety of rock excavation techniques. The study result reveals that rock strength variation affects both the blast outcome (BPR) and the quantity of explosives used in each blast round.
S. Alamdari; M.H. Basiri; A. Mousavi; A. Soofastaei
Abstract
The haul trucks consume a significant energy source in open-pit mines, where diesel fuel is widely used as the main energy source. Improving the haul truck fuel consumption can considerably decrease the operating cost of mining, and more importantly, reduce the pollutants and greenhouse gas emissions. ...
Read More
The haul trucks consume a significant energy source in open-pit mines, where diesel fuel is widely used as the main energy source. Improving the haul truck fuel consumption can considerably decrease the operating cost of mining, and more importantly, reduce the pollutants and greenhouse gas emissions. This work aims to model and evaluate the diesel fuel consumption of the mining haul trucks. The machine learning techniques including multiple linear regression, random forest, artificial neural network, support vector machine, and kernel nearest neighbor are implemented and investigated in order to predict the haul truck fuel consumption based on the independent variables such as the payload, total resistance, and actual speed. The prediction models are built on the actual dataset collected from an Iron ore open-pit mine located in the Yazd province, Iran. In order to evaluate the goodness of the predicted models, the coefficient of determination, mean square error, and mean absolute error are investigated. The results obtained demonstrate that the artificial neural network has the highest accuracy compared to the other models (coefficient of determination = 0.903, mean square error = 489.173, and mean absolute error = 13.440). In contrast, the multiple linear regression exhibits the worst result in all statistical metrics. Finally, a sensitivity analysis is used to evaluate the significance of the independent variables.
Muhammad Ahsan M.; T. Celik; B. Genc
Abstract
The distribution of stream sediments is usually considered as an important and very useful tool for the early-stage exploration of mineralization at the regional scale. The collection of stream samples is not only time-consuming but also very costly. However, the advancements in space remote sensing ...
Read More
The distribution of stream sediments is usually considered as an important and very useful tool for the early-stage exploration of mineralization at the regional scale. The collection of stream samples is not only time-consuming but also very costly. However, the advancements in space remote sensing has made it a suitable alternative for mapping of the geochemical elements using satellite spectral reflectance. In this research work, 407 surface stream sediment samples of the zinc (Zn) and lead (Pb) elements are collected from Central Wales. Five machine learning models, namely the Support Vector Regression (SVR), Generalized Linear Model (GLM), Deep Neural Network (DNN), Decision Tree (DT), and Random Forest (RF) regression, are applied for prediction of the Zn and Pb concentrations using the Sentinel-2 satellite multi-spectral images. The results obtained based on the 10 m spatial resolution show that Zn is best predicted with RF with significant R2 values of 0.74 (p < 0.01) and 0.7 (p < 0.01) during training and testing. However, for Pb, the best prediction is made by SVR with significant R2 values of 0.72 (p < 0.01) and 0.64 (p < 0.01) for training and testing, respectively. Overall, the performance of SVR and RF outperforms the other machine learning models with the highest testing R2 values.
Exploration
N. Mahvash Mohammadi; A. Hezarkhani
Abstract
Identification and mapping of the significant alterations are the main objectives of the exploration geochemical surveys. The field study is time-consuming and costly to produce the classified maps. Therefore, the processing of remotely sensed data, which provide timely and multi-band (multi-layer) data, ...
Read More
Identification and mapping of the significant alterations are the main objectives of the exploration geochemical surveys. The field study is time-consuming and costly to produce the classified maps. Therefore, the processing of remotely sensed data, which provide timely and multi-band (multi-layer) data, can be substituted for the field study. In this study, the ASTER imagery is used for alteration classification by applying two new methods of machine learning, including Random Forest and Support Vector Machine. The 14 band ASTER and 19 derivative data layers extracted from ASTER including band ratio and PC imagery, are used as training datasets for improving the results. Comparison of analytical results achieved from the two mentioned methods confirmed that the SVM model has sufficient accuracy and more powerful performance than RF model for alteration classification in the study area.
