-
Research Article
A Comparative Study on the Application of Intelligent Models in the Estimation of Backbreak in Mine Blasting Operations
Festus Kunkyin-Saadaari*,
Victor Kwaku Agadzie,
Richard Gyebuni
Issue:
Volume 9, Issue 1, April 2024
Pages:
1-13
Received:
7 December 2023
Accepted:
4 January 2024
Published:
18 January 2024
Abstract: Backbreak in the mining industry presents a considerable challenge, impacting both safety and operational efficiency. Accurate prediction of backbreak is therefore a critical endeavour. This study rigorously evaluates four advanced machine learning (ML) techniques—Lagrangian Support Vector Machine (LSVM), Radial Basis Function Neural Network (RBFNN), Gaussian Process Regression (GPR), and Extreme Gradient Boosting (XGBoost)—to ascertain the most effective method for backbreak prediction. Utilising a comprehensive dataset of 60 blasting rounds from the Damang Goldfields Open Pit Mine and prior to the analysis, this dataset underwent a thorough preprocessing phase. The efficacy of each model is assessed using a suite of metrics, including correlation coefficient (r), coefficient of determination (R2), mean squared error (MSE), root mean squared error (RMSE), and mean absolute error (MAE). The performance of the models is quantitatively compared, revealing XGBoost as the superior predictor in this context, characterised by an r of 0.9788, an R2 of 0.9565, an MSE of 0.1714, an RMSE of 0.4139, and an MAE of 0.2819. The findings of this study underscore the potential of XGBoost as a robust tool for backbreak prediction, offering mining companies a viable solution to enhance safety protocols and mitigate financial losses related to backbreak incidents. This research contributes significantly to the field of predictive analytics in mining, providing a comprehensive comparative analysis of various ML techniques for backbreak prediction.
Abstract: Backbreak in the mining industry presents a considerable challenge, impacting both safety and operational efficiency. Accurate prediction of backbreak is therefore a critical endeavour. This study rigorously evaluates four advanced machine learning (ML) techniques—Lagrangian Support Vector Machine (LSVM), Radial Basis Function Neural Network (RBFNN...
Show More
-
Research Article
S-Box Generation Algorithm by Constructing the Non-Singular Adjacency Matrix Using the Genetic Algorithm
Issue:
Volume 9, Issue 1, April 2024
Pages:
14-20
Received:
20 December 2023
Accepted:
4 January 2024
Published:
18 January 2024
Abstract: In today's applications of block ciphers, the substitution box (S-box) serves as a critical nonlinear component that is essential for generating complex ciphertext. S-boxes that exhibit lower differential uniformity and increased nonlinearity are more adept at resisting cryptanalytic efforts. The paper proposes that the construction of an 8x8 S-box can be accomplished by selecting non-singular adjacency matrices derived from graph parameters generated by a genetic algorithm. This selection is followed by an affine transformation. This method uses any graph with 8 vertices and its edge count, resulting in a non-singular adjacency matrix. The S-box is then generated by an affine mapping technique using the non-singular adjacency matrix, similar to the approach of the Rijndael algorithm. The effectiveness and reliability of the resulting S-box was rigorously tested against various cryptographic standards. The robustness evaluation included factors such as non-linearity, differential approximation probability, linear approximation probability and strict avalanche criteria. A thorough investigation confirmed that the newly created S-box met the required algebraic properties. Furthermore, a comparative analysis was performed to evaluate the performance of this novel S-box against the most recent counterparts in the literature. In terms of defense against potential malicious exploits, the results indicate a significant advantage. Overall, the results of this study underscore the significant promise and advantages of the proposed S-box-centric cryptographic strategy, positioning it as an attractive alternative to conventional encryption techniques.
Abstract: In today's applications of block ciphers, the substitution box (S-box) serves as a critical nonlinear component that is essential for generating complex ciphertext. S-boxes that exhibit lower differential uniformity and increased nonlinearity are more adept at resisting cryptanalytic efforts. The paper proposes that the construction of an 8x8 S-box...
Show More
-
Research Article
Development of Process Safety Cumulative Risk Assessment and Visualization Model/Framework for Petroleum Facilities in Niger-Delta Region, Nigeria
Emeka Maduabuchi*,
Gogomary Oyet Israel
Issue:
Volume 9, Issue 1, March 2024
Pages:
21-31
Received:
23 February 2024
Accepted:
6 March 2024
Published:
19 March 2024
DOI:
10.11648/j.ajset.20240901.13
Downloads:
Views:
Abstract: One of the key challenges in preventing major process safety accidents in an operating plant is the lack of an integrated system/model that brings together the risks posed by the deficiencies / deviations on the safety critical barriers, for operational decision making. Based on this context, an exploratory study was undertaken to develop a model/framework for visualizing the accumulation of process safety risks arising from safety critical barriers impairments in petroleum facilities in Niger-Delta Nigeria. A “focused group” was used to test/validate the model/framework using two case studies. The results indicate that the process safety cumulative risk assessment framework/model offers a transparent mechanism for assessing and visualizing the cumulative risks arising from the barrier impairment problems. For the facility in the first case study, 3.2% of the total number of safety-critical barriers was deviated and the model revealed risk accumulation in the gas compression functional location. For the facility in the second case study, 1.7% of the total number of safety-critical barriers was deviated and the model revealed risk accumulation in the gas dehydration functional location. When applied properly, the model/framework will reduce the risk of major accident in petroleum facilities by (a) aiding better management of safety critical barriers deviations through improved risks visual and (b) eliminate variability in human interpretation of process safety risk levels. One improvement area identified in the model/framework is the need for a web-based software for automation of barrier impairment data collection and real-time visualization of the cumulative risk picture.
Abstract: One of the key challenges in preventing major process safety accidents in an operating plant is the lack of an integrated system/model that brings together the risks posed by the deficiencies / deviations on the safety critical barriers, for operational decision making. Based on this context, an exploratory study was undertaken to develop a model/f...
Show More
-
Research Article
Investigating the Effect of Palm Kernel Shell Powder on the Rheological and Filtration Properties of Water Based Mud
Issue:
Volume 9, Issue 1, March 2024
Pages:
32-41
Received:
20 February 2024
Accepted:
15 March 2024
Published:
2 April 2024
Abstract: During drilling operations, the use of drilling fluid plays a critical role, and over time, there has been considerable interest in enhancing drilling fluid characteristics in order to improve performance, reduce costs, and prevent environmental pollution. Deviating from conventional additives, recent studies have explored the use of alternative materials, as drilling fluid additives. In line with this trend, this study focuses on the laboratory investigation of the rheological and filtration properties of water-based drilling fluid treated with Palm Kernel Shell Powder (PKSP) with high viscosity polyanionic cellulose (PAC HV), used as control. To assess the impact of PKSP in water-based mud, experiments were carried out using concentrations spanning from 0.5g to 2.5g, temperatures of 27°C, 40°C, 60°C, and 80°C, and aging of 24, 48, and 72 hours. From the results the plastic viscosity of mud samples treated with PKSP were temperature dependent and also with increasing aging. The addition of PKSP showed improved performance in terms of reducing the filtrate volume as well as the cake thickness with increasing concentration of the additives, and the concentration that gave the best results across all aging duration was 2.5g. The mud weight and pH of all samples remained relatively constant, with no significant changes observed. However, PAC HV showed better results in all the cases of fluid loss and mud cake thickness. It could be attributed to the soluble contents in the PAC HV which increased the viscosity significantly and thus, kept the solid particles in suspension.
Abstract: During drilling operations, the use of drilling fluid plays a critical role, and over time, there has been considerable interest in enhancing drilling fluid characteristics in order to improve performance, reduce costs, and prevent environmental pollution. Deviating from conventional additives, recent studies have explored the use of alternative ma...
Show More