AP26198535 – Ensuring safe underground mining of precious metals based on digital modeling a network of intersecting excavations and big data analysis

Objective of the project – to identify patterns of formation of zones of catastrophic rock pressure during underground mining of precious metals (gold), which will ensure the stability of rocks and the safety of mining operations, especially in areas of geological disturbances.

Relevance: The relevance of this project is driven by the need to enhance the safety of mining operations in complex geological and tectonic conditions of Kazakhstan’s deposits. Effective prediction of rock mass behavior requires a comprehensive approach that includes accounting for the natural stress field and its changes under the influence of technological processes. The determination of the tectonic component of the stress state is of particular importance, as it significantly affects deformation and failure of the rock mass. Insufficient accuracy in assessing the strength and deformation properties of rocks may lead to modeling errors and an increased risk of emergency situations. The application of modern digital technologies and numerical modeling methods allows for more accurate identification of potential failure zones and assessment of the stability of mine workings. The development of intelligent forecasting systems based on machine learning will enable timely prevention of hazardous rock pressure manifestations and improve the reliability of mining processes.

Scientific supervisor: Doctor of technical sciences, Professor, Moldabayev Serik

Expected and achieved results: Within the framework of the project, the natural triaxial stress field in the Zholymbet deposit area, including zones of potential geological disturbances, was determined based on ISRM recommendations. Field measurement stations were established at depths of 720 m and 820 m, where stress measurements were carried out using the hydraulic fracturing method. A total of 35 hydraulic fracturing tests were performed, which revealed a significant tectonic component in the stress state of the rock mass. The azimuth of the maximum horizontal stress was determined to be 40° ± 15°, and the ratios of principal stresses were obtained for subsequent geomechanical modeling. Methods for sampling and determining the physical and mechanical properties of rocks were developed and applied in an accredited geomechanics and geotechnology laboratory. Uniaxial compressive strength (UCS) tests were conducted to determine strength and deformation characteristics, including Young’s modulus and Poisson’s ratio. Data on ultrasonic wave velocities were obtained, allowing additional assessment of the rock mass condition. An analysis of intelligent data processing methods was carried out, and appropriate machine learning approaches were selected for classifying rock mass sections by hazard level. Requirements for training datasets were formed, including the need for proper labeling and consideration of limitations related to incomplete data. The applicability of unsupervised and semi-supervised learning methods was justified for identifying hidden patterns in geomechanical data. A hybrid approach combining numerical modeling in the RS3 environment and machine learning techniques was developed. It was shown that integrating modeling results with data analysis algorithms improves prediction accuracy and ensures interpretability in assessing the stability of underground workings.

List of publications with links to them

1. Молдабаев С.К., Асылханова Г.Н., Асылханова С.А. Секторальное моделирование открытых горных выработок в программе RS3 компании Rocscience // Горный журнал Казахстана. – Алматы, 2025. – № 8. – С. 12–16. DOI: https://doi.org/10.48498/minmag.2025.244.8.009 – ISSN 2227-4766.
2. Молдабаев С.К., Асылханова Г.Н., Турсбеков С.В., Асылханова С.А. Влияние угла наклона бортов глубокого карьера на полноту безопасного извлечения запасов месторождения // Университет Еңбектері / Труды Университета. – Караганда, 2025. – № 4. – (в печати). – ISSN 1609-1825.
3. Молдабаев С.К., Султанбекова Ж.Ж., Молдабаев А.С., Моисеев В.А.
   Секторальное моделирование в программе RS3 компании Rocscience и учет тектонических разломов : авторское свидетельство № 55695. – Казахстан, 2025.