The paper examines the possibilities of applying big geospatial data analysis and machine learning methods to assess the anthropogenic impact of mining activities on natural ecosystems. Multispectral satellite imagery from the Sentinel-2 remote sensing system was used as the primary data source. The preprocessing stage included georeferencing, atmospheric correction, and calculation of the NDVI vegetation index. A Random Forest machine learning algorithm was applied to automatically identify technogenically disturbed areas. The analysis revealed significant differences in spectral characteristics between natural landscapes and mining-affected territories. The classification model demonstrated an overall accuracy of 87.2%, confirming the effectiveness of machine learning techniques in processing geospatial information. The results highlight the potential of integrating satellite remote sensing data with intelligent analytical methods for environmental monitoring and for assessing the spatial distribution of anthropogenic transformations in mining regions.
The paper addresses the problem of automated analysis of large industrial monitoring datasets for identifying early signs of pre-emergency conditions in mining systems. An approach for processing time series of technological parameters is proposed, including data cleaning, normalization, and the application of intelligent data analysis methods. The study is based on archival monitoring data of a crushing and conveyor complex, including measurements of temperature, vibration, and mechanical load parameters. The Isolation Forest algorithm was used to detect anomalous operating modes and was trained on a dataset representing normal equipment operation. The analysis results demonstrated the ability of the proposed method to automatically identify deviations in technological parameters that may indicate the formation of potentially hazardous operating conditions. The findings confirm the potential of machine learning techniques for improving the effectiveness of industrial monitoring systems and enhancing safety in mining equipment operation.
The paper addresses the problem of predicting geoecological risks in mining territories using machine learning techniques. The relevance of the study is associated with increasing anthropogenic pressure on the environment in regions of intensive mineral resource development and the need for effective tools for environmental risk assessment. An intelligent system for geoecological risk prediction based on a hybrid ensemble approach combining Random Forest, Gradient Boosting, and k-Nearest Neighbors algorithms is proposed. Experimental data obtained from environmental monitoring, including indicators of water and soil contamination, hydrogeological parameters, and anthropogenic load indices, were processed and analyzed. The results of experimental modeling demonstrate that the ensemble model improves prediction performance, achieving a classification accuracy of 0.89 and reducing prediction errors compared to individual algorithms. The developed approach shows high potential for application in environmental monitoring systems and in the assessment of geoecological risks in mining regions.
The paper considers the possibility of applying machine learning algorithms for diagnosing the technical condition of industrial equipment at mineral resource industry enterprises. The relevance of the study is determined by the high complexity of technological processes and the need to improve the reliability of industrial systems operation. Traditional approaches to technical diagnostics based on scheduled inspections and analysis of equipment operating parameters are analyzed. It is shown that modern data analysis methods allow processing large volumes of information obtained from sensors and industrial control systems, identifying anomalies in equipment operation and predicting possible failures. The feasibility of implementing machine learning algorithms in industrial monitoring systems is substantiated. The obtained results demonstrate the prospects of using intelligent data analysis methods to improve maintenance efficiency and reduce the risk of emergency situations in industrial enterprises.
The article presents the development of a methodological approach for identifying and quantitatively assessing technogenic hazards occurring in technological processes of mineral extraction and primary processing. The study is based on the analysis of risk factors typical for mining enterprises and on the application of an integrated hazard index that considers the probability of hazardous events and the severity of their potential consequences. The research focuses on key production operations including drilling and blasting, transportation of rock mass, and primary crushing processes. Based on the analysis of technological operations, the main technogenic risk factors were identified and quantitatively evaluated using statistical production data and expert assessments. The modeling results made it possible to determine the most significant sources of risk and to perform a comparative assessment of hazard levels for different technological operations. The proposed approach can be used to improve industrial safety management systems in mining enterprises.
The paper examines the features of pollutant migration in soil and ground systems of mining industrial territories using ecological and geochemical modeling. The aim of the study was to analyze the patterns of redistribution of contaminating components under conditions of technogenically transformed geological environments. To achieve this goal, a computational model of a vertical soil and ground profile was developed, including a surface soil horizon, technogenically altered soils, and underlying natural rocks. Numerical simulation of mass transfer processes was carried out using the finite difference method in the MATLAB environment, taking into account both diffusion and infiltration transport mechanisms. The modeling results demonstrated the formation of a secondary accumulation zone of pollutants at depths of approximately 0.6-1.0 m and the gradual downward movement of the contamination front. The obtained results make it possible to assess the dynamics of pollutant redistribution and can be applied in environmental assessment and forecasting of the ecological state of mining industrial territories.
This study assesses the stability of large-section stopes used for mining thick, steeply dipping ore bodies using caving systems, taking into account the structural and petrographic features of the massif. Based on a combination of studies, including stereophotogrammetric fracturing analysis and petrographic studies, a geomechanical model was developed that describes the distribution of stress fields during the mining of a blind ore deposit. It was established that, despite extensive fracturing, the presence of high-strength mineral inclusions in the ores and rocks contributes to the massif's ability to accumulate significant elastic deformations, increasing the risk of sudden failures. The modeling results indicate a relatively stable state of the mined-out space; however, zones of stress concentration were identified that require monitoring during mining.
The formation peculiarity of ceiling pillars between the Gremyachinskoedeposit workings is that the pillars have a complex spatial configuration. The paper substantiates an engineering approach for determining the estimated degree loading of the ceiling beams in the deposit conditions. Based on the numerical experiment results, the concentration coefficient of horizontal stresses acting on the ceiling from the host rocks was determined. The influence of the ceiling beam geometry on its load-bearing capacity has been established. Based on the results obtained, the minimum allowable ceiling thickness was estimated in typical conditions of a deep potash mine. The results of the work are intendedto assess the permissible thickness of ceilings in the Gremyachinskoefield.
The article examines the issues of modeling structural changes in a regional economy under the conditions of climate and industrial policy implementation. The relevance of the study is determined by the growing global climate agenda and the need to adapt regional economic systems to new institutional and technological development conditions. The purpose of the research is to develop and test an approach for analyzing possible transformations in the sectoral structure of a region using economic and mathematical modeling of interindustry relations. The methodological basis of the study is a simplified interindustry model that makes it possible to assess the influence of changes in investment activity and production dynamics in individual sectors on the structure of gross regional product. The conducted computational experiments made it possible to identify the most sensitive elements of the regional economic system and determine potential directions of its structural transformation. The results obtained can be applied in the development of regional economic policy and in designing scenarios for sustainable socio-economic development.
As a result of the conducted field geological and geomorphological research, the application of stock and cartographic material, as well as the interpretation of aerospace images of 2000-2025 at a scale of 1:60000, the territory of the Gobustan was divided into zones according to the distribution of geological processes and relief forms: mid-mountains, lowlands and lowland zone, in which a number of geological processes pose a real threat to the development of recreational and tourist activities that are exclusively dependent on the relief, and the construction of the corresponding infrastructure.
The article examines the socioeconomic consequences of the global energy transition for resource-oriented regions. The strengthening of international climate policy and the rapid development of low-carbon technologies are transforming the structure of the global energy system and changing the conditions for territories specialized in the extraction and processing of natural resources. The aim of the study is to analyze the key factors and mechanisms influencing the socioeconomic development of such regions under the conditions of the energy transition. The paper identifies major risks associated with declining demand for traditional fuels, shifts in investment flows, and potential structural changes in regional labor markets. At the same time, the study highlights development opportunities related to economic diversification, technological modernization of industries, and the stimulation of innovation activity. The findings can be applied in the development of regional economic and energy policies.
The paper describes a technique for rationalizing the unit volume of an explosive unit in the production of special explosive work under cover, the essence of which is the use of special separate shelters-mats made of non-woven fabric. This makes it possible to optimize the volume of a mass explosion due to the possibility of blocking the deterrent factor- an increase in the unit volume of the explosive block, due to the requirement of clause 180 of the Federal Norms and Rules of the & laquo;Rules for Handling Explosive Materials & raquo;, allowing the detonation of a group of charges covered with protective devices only if they explode simultaneously or with a total deceleration of no more than 200 ms. The above factor of differentiation of the surface shelters of the explosive block, combined with drilling inclined wells (at an angle of 750 relative to the daytime surface, directed in the opposite direction from the protected object) , the use of an explosive network installation scheme with an increase in the specific inter-well deceleration towards the protected object, makes it possible to increase the volume of a mass explosion by two or more times relative to special explosions with the use of continuous shelters. The introduction of an appropriate technique at a local facility in specific mining, geological and technological conditions is accompanied by control measurements of seismic vibrations for each step of increasing the unit volume of the explosive block, with the recording of the ground displacement rate and the amplitude of vibrations. This technique has been suc- cessfully tested at the Koksovy coal mine in the Kemerovo region.
Positive trends in the change of environmental land conditions in the Tula Region have been identified for the period from 2001 to 2020. The shares of lands classified as degraded, stable, or improved over time have been determined. Non-degraded lands dominate the composition of the land fund. A statistically reliable non-linear correlation has been established between the indicator of land's ecological condition, specific emissions of polluting substances from stationary sources, and population density values in the administrative units of the Tula Region. For 38 % of the territories, there is an increase in the risk of land degradation associated with higher levels of emissions.The urban district of Novomoskovsk, the village of Novogurovsky, as well as Uznovskiy, Efremovskiy, Kimovskiy, Volovskiy, and Venevskiy districts are characterized by high proportions of degraded lands in their land structure accompanied by high rates of pollutant emissions.
The article addresses the problem of detecting pre-accident states of mining engineering systems using industrial monitoring data. Modern mining enterprises are characterized by complex technological processes and a high level of operational risks, which requires efficient approaches to processing monitoring information. The study proposes an intelligent data analysis system based on machine learning methods for anomaly detection in multivariate time series of technological parameters. The system includes data preprocessing procedures and the Isolation Forest algorithm, which allows identifying abnormal deviations in monitored parameters. Experimental testing was carried out using archived monitoring data from a mining enterprise. The obtained results demonstrate the possibility of automatically detecting potentially hazardous operating conditions of equipment and confirm the effectiveness of applying intelligent data analysis methods to improve industrial safety and operational reliability in mining systems.
The article considers the application of interpretable machine learning models for assessing the environmental and industrial safety of mining enterprises. A regression decision tree algorithm was used as the main methodological approach, enabling not only prediction of the integrated safety indicator but also interpretation of the influence of individual production and environmental factors. The study is based on statistical data describing technological parameters of mining operations, pollutant emissions, waste generation, energy consumption, and equipment reliability. The experimental procedure included data preprocessing and normalization, model training, and evaluation of predictive accuracy. Additionally, feature importance analysis was carried out using the SHAP method. The results show that specific emissions, energy consumption, and equipment reliability have the greatest influence on the safety indicator. The proposed approach can be used as a tool for analytical support of environmental and industrial risk management in mining enterprises.
The article considers stochastic assessment of the remaining useful life of mining equipment under variable operating loads and incomplete information on wear processes. A degradation-based approach is proposed that uses an integral technical condition indicator and a probabilistic description of reaching the limit state. A stochastic model of remaining useful life is developed, the principles of its parameter identification from operational data are discussed, and the applicability of the obtained estimates to maintenance and repair planning is demonstrated.
The paper considers the problem of improving the efficiency of ventilation and local dust suppression systems in underground mine workings as an important factor in ensuring industrial safety at mining enterprises. The aerological conditions of aAmine working were analyzed and experimental studies were carried out to assess the influence of airflow parameters and water spraying on dust aerosol concentration in the working area. Measurements were performed under different ventilation regimes and water pressure levels in the spraying system. The results show that the use of ventilation or dust suppression alone provides only a limited reduction in dust concentration. The highest efficiency is achieved when these methods are applied simultaneously. Experimental results demonstrate that increasing the airflow velocity to about1 m/s combined with water spraying at a pressure of 0.6 MPa reduces dust concentration by 68-71% compared with the initial conditions and significantly shortens the air purification time in the working area.
The article addresses the problem of assessing investment risks and financial stability of companies in the mining sector. A methodological approach for an integral economic assessment based on the analysis of key financial indicators such as liquidity, capital structure, profitability and business activity is proposed. The study uses financial statements of mining enterprises for several reporting periods as the data base. The indicators were normalized relative to industry averages and an integral index of financial stability was calculated. The proposed methodology was tested on three mining companies, which made it possible to identify differences in their levels of financial sustainability and investment risk. The results demonstrate that the approach can be applied for comparative analysis of companies and rapid assessment of investment attractiveness in the resource sector of the economy.
The article addresses the problem of intelligent anomaly detection in industrial monitoring data for assessing the stability of mining engineering systems. Modern mining enterprises generate large volumes of monitoring data that reflect the state of technological processes and engineering structures. Traditional approaches to data analysis often have limitations related to insufficient sensitivity to complex parameter relationships and difficulties in processing large datasets. The study proposes an approach to the intelligent analysis of industrial monitoring data based on machine learning methods. The Isolation Forest algorithm is used as the main tool for anomaly detection, allowing efficient identification of atypical observations in multidimensional time series. The proposed method was tested on real industrial monitoring data. The results demonstrate its ability to detect abnormal operating conditions and provide informative indicators for assessing the operational stability of mining systems.
The issues of improving environmental protection and resource-saving technologies for the development of ore deposits by the underground method with the filling of the worked-out space with concrete based on waste from mining and processing of ores are considered. The research isbased on the idea of optimizing concrete compositions by using industrial waste as binders.
