Home – Home – Jornals – Advanced Search

The paper presents a comparative analysis of various machine learning algorithms for solving the problem of predicting the heat transfer crisis during boiling in two-phase flows inside channels of various geometries. Twelve classical regression models implemented in the Scikit-learn, LightGBM, XGBoost, and CatBoost libraries, as well as neural network methods are considered. The models are compared with each other, as well as with traditional forecasting methods based on the use of skeletal tables, approximate semi-empirical ratios, and correlation formulas. Possibilities of hybrid models that combine the approach based on domain knowledge with machine learning algorithms are discussed. The results of experiments with a model that combines the CatBoost regressor with one of the traditional methods in a hybrid scheme are presented. The advantage of machine learning models over the traditional approaches is revealed. It is shown that the best performance for all metrics among machine learning models can be achieved by using ensembles of algorithms based on gradient boosting.

The author presents a 3D structured packing with a coordination index equal to 8 to replace effectively a real-life random packing of particles. The notions of vectors of principal stresses and plastic strain rates are introduced. 3D equations of plastic deformation are constructed. They fulfill the model adequacy requirement-zero energy dissipation in the medium of perfectly smooth particles. In the presence of internal friction, the equations lead to a nonassociated flow rule.

The stress-strain behavior patterns in rockburst-hazardous rock mass are described as a case-study of mineral deposits in Gornaya Shoria. The microseismic research data on development of rockburst hazard criteria are described. Geodynamic events are predicted in extraction blocks during basting, with detection of possible damages in underground openings using the obtained patterns of different energy shocks.

Based on the mathematical modeling of the geomechanical condition of a coal face, it is shown that destruction of the face zone occurs in the form of squeezing-out and spalling within an operational cycle. The acoustic noise in the longwall is analyzed, and the noise spectrum peak frequency is determined. The peak frequency is reflective of the size of the damage area in the face zone.

The paper concerns an interpretation of high-precision gravimeter observations in the Upper Kama Potash Salt Deposit. The studies aim to detect the mining-induced softened areas in undermined rock mass. It is found that the gravity field transformants exhibit induced softening areas as flat inclined negative anomalies that intersect the undermined rock mass from top downward. The spatial location of the induced softening areas in rock mass is determined.

For this study, as part of the underground operations, the underground mining sequence was explored for suitability in the Zuuntsagaan mine. Two mining sequences are considered; bottom-up (B-U) and top-bottom (T-B), and the instability of stopes and open pit slopes is evaluated in response to each sequence. For the present rock mass conditions and stress state, with reference to the underground section, the results revealed that T-B sequence has lower ground movements. On the open pit section, the effect of the two sequences is relatively the same while the difference is mainly observed in the underground section.

In this study, a path search algorithm is introduced for exploring the ultimate pit limit using the Lerchs-Grossmann method. In an iteration of the algorithm, each vertex of the solution is randomly moved within a specified neighborhood. The vertices’ motion in such a neighborhood ensures that the formation of abnormal shapes is prevented. Since an open-pit design needs to address geomechanical concerns coincident with economic value maximization, the movements resulting in segments steeper than an allowable slope angle are discarded. For each valid vertex movement, the benefit of the resulted pit limit was calculated. The new position of the vertex is accepted according to a probabilistic regime in order to prevent the algorithm be trapped in a local maximum. Independence of the slope from the block size is one of the main advantages of the suggested algorithm. Hence, the algorithm provides a better mapping to the ore body and results in a higher benefit. The reason is the mobility of the path search algorithm and its high flexibility.

The authors discuss the approved advanced control charts for operational cycles of hydraulic impacting machines. These control charts allow adjusting operational characteristics of the machines either using only pressure back-coupling between the fluid power system and distributor, or in combination with introduced additional pressure control. The article describes adaptive impacting machines equipped with the control charts which allow adjustment of blow energy and frequency subject to properties of a medium being fractured. It is emphasized that to expand the application range of hydraulic impactors, it is required to design high-frequency percussion machines with high impact capacity owing to improvement of distributors and due to transition to the increased power fluid pressure.

The author analyzes spectra of a downhole impulsive vibration source with two power-supply units and an electromagnetic hammer using the Fourier series. The amplitude-frequency characteristics of the impulsive and unbalance-mass vibration sources of adequate sizes are compared. The impulsive vibrator has much higher amplitudes and spectra of output signals than the unbalance-mass vibration source.

The study addresses efficiency of natural cold and the permafrost properties in the Central and Arctic Yakutia in holding of frozen cola in buried storage rooms. The heat transfer modeling takes into account the climate, parameters of a storage room and heat insulation. It is shown that coal thaws less than the host rocks because of its low thermal conductivity. When a storage room is filled in winter, coal will remain frozen for a few years, and heat insulation will greatly reduce the rate of thawing in the overburden. It is emphasized that as against the width and slope of the storage room, its occurrence depth is the main factor to govern the size of the thawed zone by the end of the warm period of storage. Natural cold in the buried storage room decreases coal oxidation, preserves coal properties, and improves energy security of the hard-to-reach areas.