Problems of mathematical modeling of three-phase flow of immiscible fluids: oil, gas and water, are considered. The system of differential conservation equations is associated with a completely explicit numerical scheme on a super-element grid. A technique is proposed for producing wells working at bottomhole (and reservoir) pressure below the bubble point pressure in the superelement model (for large cell sizes and a significant pressure change inside the cell). Satisfactory results were obtained in comparative testing using commercial software.

The mathematical model of components separation of a dispersed medium in a spatially non-uniform flow of merging particles in collisions is considered. The availability of the nonuniform velocity field of particles transport generates a new effect connected to the analog of cumulative flows where space-time zones are formed with an intensive release of the polymeric structure that is instantaneously removed from the medium (sedimentation based on the separation of the precipitated polymer). At the same time, space-time features of functions that describe the concentrations of particles in a dispersed medium arise in the mathematical description of the phenomenon. A computational algorithm for the approximations concentration constructing is proposed, test calculations are performed, and the Monte Carlo method with duplicate tests is compared with calculations of explicit difference scheme. A data preparation procedure of data preparation at the level of individual particles is proposed, leading in the limit to a solution of a convergent difference scheme, which in turn converges to a generalized solution of the Cauchy problem for the spatially nonuniform Smoluchowski equation of kinetic theory of coagulation.

The statistical instability of successive samplings of the same biological system in continuing homeostasis is studied within the statistical approach. It is shown that matrices of pairwise comparisons of parameters samplings x_i of biosystem state vector x = x(t) = (x₁, x₂, ..., x_m)^T demonstrate the instability of statistical functions distribution f(x) (i.e., f_j(x_i) = f_{j + 1}(x_i)). At the same the statistical stability is demonstrated by the number k and statistically matched pairs of samples x_i. The probability P of obtaining two coincident successive samplings x_i is extremely low (P ≤ 0.05) for any repetitive processes of homeostatic systems.

The article proposes a technique for the structure and parameter identification of systems described by models of single-phase and two-phase filtration in a porous medium with the presence of strong heterogeneities. There are some formulated principles that need to be followed during constructing the model structure. The paper also includes an algorithm for detecting fractures and sealing faults using only production data. A good agreement of the results obtained on the reference model and on the model constructed with the use of the proposed structure and parameter identification technique is shown.

The model extension of nonlinear preferential attachment random graphs considering possibilities of adding entire communities in the network is proposed. The added communities are represented by entire graphs. Such an extension of model is naturally based on existing processes on social networks. Mathematical relations are derived. These relations allow calculating the degree distribution of vertices {Q_k} of a generated graph. Equations are also derived which allow, for a given node degree distribution of real network getting generation properties considering the process of adding communities. The results obtained are verified by generating large graphs.

It is necessary to simulate the events that occur during cloud flying when modeling outside world in flight simulators. In addition to that, visualization system should work in real-time with no less than 25 frames per second. This article presents an analysis of different types of cloud visualization: procedural modelling, modelling with the use of cellular automaton, skyboxes, the Perlin noise used in conjunction with billboard and voxel rendering. Major disadvantages of these methods are presented, such as computational complexity, lack of realism and inability to fly through the clouds. It is demonstrated that the use of the Perlin noise in conjunction with billboards provides better performance of the visualization system and more realistic results.

There are various tools currently helping to easily alter files. This paper considers genuineness identification problems of photo, video and audio files.

The paper presents the development of an automated calculation system of tolerances screening for electronic components of technical devices. The study is based on an analysis and implementation for the layout of organizational-functional model of the system, physical model of the database and developed user interface.

The article discusses standard theoretical conceptions, relevance and practical significance of machine learning in the oil and gas industry. The article provides information on the benefits of using the machine learning.

The article examines electrical networks parameters (6-110) kV of oil production, in the sphere of innovative development of the power industry in developing regions of Russia. The functional and technological ideology of the Smart Grid concept and methodological base for researching electric power systems cybernetics are examined to determine their identification in terms of specific tasks. Conclusions are drawn.

The paper studies the choice of calculation method of liquid-gas mixture flow for use in decision making systems in well oil production. Calculations were made using three methods and conclusions about their applicability in information systems of geophysicist decision making were drawn.

The paper proposes new multi-texturing methods and algorithms for objects of 3D virtual environment. The task of object rendering with blend materials is considered. Such materials consist of several sub-materials structured in the form of a binary tree and laid on the same area of the object surface. To blend results obtained by independent application of each material from the tree to the object’s surface, the visualization process uses per-pixel mixing. Proposed solutions are based on the use of distributed computing by means of modern multi-core GPUs.

The paper presents the results of a study of biomedical data, as well as data obtained from receivers using mathematical statistics in the oil and gas industry. As biomedical data, the RR intervals were used for cardiac rhythm, and the parameters of water pipe sensors were used as characteristic data. The study determined the correlation dependence between progressive levels of the time series with time-shifted data series. An analysis of the autocorrelation function and the correlogram is carried out.

The article considers a construction of regional development trajectories and working of their classifier. The authors have proposed a resource model supplemented by the J. von Neumann model of expanding economy that is built for five subsystems selected by functional differences: economic, social, infrastructural, natural resources and external. The method of constructing regional development trajectories and possible approaches to class structuring for these trajectories, based on magistral trajectories, was proposed. The authors have defined the possibility of comparing an in-place regional trajectory and a documented one.

The article considers an data transmision system upgrade. It is based on one of the spectrum extention methods of transmitted signal that is the frequency-hopping spread spectrum. Analysis of the data transmission systems development shows that the main trend of changes at a level of data processing and transmission is the qualitative complication of physical organization for realization of new methods of adaptive signal processing. The use of a cellular neural network with hysteresis to perform state analysis of available frequency channels is proposed to increase the security of transmitted information. The results of neural network functioning allow predicting possible occupancy of each available radio channel and the probability of jamming at the moment of information transfer at a certain carrier frequency. This information allows adjusting the pseudo-random sequence generator operation generating the frequency partitioning sequence of available wireless communication channel. The performance in the form of adaptive neuristor lines that process signals transmitted between neurons is proposed to increase efficiency of the cellular neural network functioning communication between neural elements. The neural network system is able to process information not only in neurons with hysteresis, but also in the space between neural elements, which allows modifying the behavior of a neural network depending on the dynamics of signal propagation between its elements. Further the issue of variable parameters control of the designed complex neural structure by means of the modified algorithm of diffusive iterative search is considered. Application of similar optimization strategy leads to formation of metaheuristics which is capable to perform successfully tuning and training of a neural network complex. Check of network ability to predict frequency intervals with the minimal interference has shown that the cellular network is capable of performing reorganization of working frequency with account for features of background noise occurrence and jamming support in available frequency channels.