Работая с сайтом, я даю свое согласие на использование файлов cookie. Это необходимо для нормального функционирования сайта, показа целевой рекламы и анализа трафика. Статистика использования сайта обрабатывается системой Яндекс.Метрика

Принять

Научный журнал Моделирование, оптимизация и информационные технологииThe scientific journal Modeling, Optimization and Information Technology

Online media
issn 2310-6018

submit
a manuscript

metadata of articles for the last 2 years

Numerical methods for solving mathematical models of the temperature distribution of strips and rolls during hot rolling with interval parameters

2024. T.12. № 1. id 1528

Dabas M.R. Saraev P.V.

DOI: 10.26102/2310-6018/2024.44.1.028

The article considers the problem of temperature distribution in the strip and working rolls during hot rolling under the conditions of uncertainty of input parameters. The zone of the deformation gap with the formation of a rolling scale strip on the surface is regarded, as a result of which a system of thermal conductivity equations with different initial and boundary conditions is solved in the area of the deformation gap being studied. Next, the zone of the interstand gap is considered, where the heat exchange of the strip with the environment occurs. In all zones, the input parameters are represented as interval numbers. The deformation gap and the interstand gap were discretized from a continuous region into a grid one, systems of linear algebraic equations with tridiagonal interval coefficient matrices were derived using finite difference approximation, and a counter-run method with interval coefficients was presented to solve the obtained systems. The article considers the calculation results for 7 stands running one after another and consisting of a deformation gap and an interstand gap for the case with real input parameters and for the case with interval input parameters, calculations were performed using the developed software for both cases.

Keywords: equation of thermal conductivity, two-sided Thomas algorithm, interval arithmetic, hot rolling, finite difference approximation

Motion control along a program trajectory using a neural network

2024. T.12. № 1. id 1527

Grinyak V.M. Shutov K.S. Artemiev A.V.

DOI: 10.26102/2310-6018/2024.44.1.024

The paper examines the creation of a hardware and software prototype of an unmanned vehicle and testing its hardware and software architecture in an attempt to create a universal standard solution for this type of device. The problem of controlling a drone is considered in such a way that it is possible to flexibly switch the sources of control commands and control algorithms. For this purpose, the subsystems for generating and executing control commands are proposed to be connected via a message queue. It is makes possible to combine autonomous and manual controlled modes of operation of the drone. A method for generating control commands when an object follows a program trajectory, based on a neural network, is proposed. The input data of the network are the coordinates of the program trajectory and the current state of the object, and the output data are control actions. The paper describes the hardware and software components of an automobile-type device, the architecture of its control system, the architecture of a neural network, and possible approaches to its training. The creation of a training set using both simulated and real traffic data is discussed, which allows the self-driving device to “learn” different driving styles. The results of experiments with various training samples are presented, which demonstrate the practical applicability of the proposed control method. Attention is paid to aspects of the neural network structure, including the choice of the number of layers and neurons. The possibility of using “intermediate” points of the program trajectory to improve the properties of the object’s movement is indicated. In general, it is concluded that the use of neural networks is promising in the control of drones, in cases where combining and flexible switching of control algorithms is required.

Keywords: unmanned vehicle, vehicle control, navigation, autonomous vehicle, neural network

Finite element modeling of a metal tubing hanger seal

2024. T.12. № 1. id 1525

Timofeev E.K. Godenko A.E. Lipatov E.Y.

DOI: 10.26102/2310-6018/2024.44.1.022

The process of underwater oil and gas production is accompanied by high values of well pressure, which can reach thousands of atmospheres, while the service life of this equipment, laid down in the technical specifications, can reach 20 years or more. To ensure the safety of the subsea production process throughout the life of the equipment, it is important to pay special attention to the tightness and strength characteristics of metal seals as part of the design. In this regard, this article is aimed at identifying the degree of influence of such a geometric parameter as the depth of the recess on the characteristics of the tightness and strength of metal seals. The finite element method was modelled, using the Ansys calculation complex, of the stress-strain state of the metal seal of a tubing hanger of a standard design during installation and under the influence of well and test pressure. Finite element modeling was carried out using elastoplastic models of material deformations. To analyze the performance of a metal seal design, criteria for assessing strength and tightness are given. During the modeling, parameters characterizing the tightness and strength of the structure were determined. In order to study the degree of influence of the recess depth of the metal seal on the characteristics of tightness and strength, modeling was performed with an increase in this parameter to 80 %. The results of calculations of tightness and strength parameters with different recess depths are presented. The materials of the article are of practical value for design engineers and scientists dealing with the problems of ensuring a hermetic underwater connection using metal seals.

Keywords: subsea production system, metal seal, stress-strain state, tubing hanger, subsea Christmas tree, contact pressure

Model of selective assembly of two elements with dependence of the output parameter as a quotient of the input parameters

2024. T.12. № 1. id 1523

Filipovich O.V.

DOI: 10.26102/2310-6018/2024.44.1.027

The technological process of single-parameter selective assembly of two elements with parameters that are random variables, the values of which are determined by the finishing operations of the manufacturing processes, is considered. It is considered that the dependence between input and output parameters is nonlinear (nonlinear input-output models) and is represented in the form of quotient, and the completing rule is elementary. For a dependence of this type, expressions linking the values of tolerances (including group tolerances), limit deviations and limit values of input and output parameters are given. A method is proposed that helps to calculate group tolerances to fulfil the requirements to the accuracy of the output parameter in the whole area of its permissible values, as well as to determine the boundaries of selective groups. It is based on an iterative procedure, with each iteration consisting of sequentially executed steps. The output data of the previous iteration are the initial data for the next one. As a criterion for the end of the procedure, a given level of accuracy in calculating the average group tolerances is taken. The analytical and probabilistic model is developed, which takes into account the calculated boundaries of selective groups and helps to determine the most important indicators of selective assembly, such as the probability of formation of assembly sets, probabilities of formation of incomplete elements. An example of modelling is given, in which process indicators are determined using the developed method and model with given initial data. Prospects for further research are outlined.

Keywords: selective assembly, mathematical model, nonlinear dependence, quotient, iterative method

Triggers of motor activity measurable by near-infrared functional spectroscopy (fNIRS): a review

2024. T.12. № 2. id 1522

Samandari A.M. Afonin A.N.

DOI: 10.26102/2310-6018/2024.45.2.004

Scientific studies have differed on the interpretation of activity in the primary motor cortex of the brain. Various studies have found that the primary motor cortex is activated only during physical motor tasks. Whereas other studies have appeared that a similar measurable activity can be observed and recorded when arousing or stimulating the motor cortex when performing a mental representation of movement. Consequently, our purpose of this review was to compare the triggers of motor cortex activation during the physical execution and mental representation of the movement by recording the brain signals resulting from the stimulation by using the technique of near-infrared functional spectroscopy based on the neural interface (brain-computer interface). This research reveals differences and comparisons based on various approaches to analyze and systematically realize target triggers of motor cortex activation during training at neural interface (fNIRS). Based on the above, this review concludes by emphasising the fact that triggers of cortical activation in general and under different names cause activity that can be recorded by measuring the various changes that occur in hamoglobin concentration, in other words, that both physical task performance and similar mental representations of movement cause perceptible activity in the motor cortex. This provides the rationale for prosthetic, rehabilitation and other applications. Furthermore, this encourages future research to identify positive triggers for cortical activation to study psychological states of cognitive function and certain pathological conditions, as well as neurophysiological studies.

Keywords: near-infrared functional spectroscopy, triggers, motor cortex, brain-computer interface, physical movement, mental representation of movement

Mathematical model of a universal control system for a walking robot based on reinforcement learning methods

2024. T.12. № 1. id 1520

Kashko V.V. Oleinikova S.A.

DOI: 10.26102/2310-6018/2024.44.1.025

Modern approaches to solving the problem of controlling walking robots with rotary links are disparate algorithms built either on a ready-made locomotor program with its further adaptation or on complex kinematic-dynamic models that require extensive knowledge about the dynamics of the system and the environment, which is often unfeasible in applied problems. Also, the approaches used are strictly related to the configuration of the walking robot, which makes it impossible to use the method in applications with a different configuration (a different number and type of limbs). This article proposes a universal approach to controlling the motion of walking robots based on reinforcement learning methodology. A mathematical model of a control system based on finite discrete Markov processes in the context of reinforcement learning methods is considered. The task is set to build a universal and adaptive control system capable of searching for the optimal strategy for implementing a locomotor program in a previously unknown environment through continuous interaction. The results distinguished by scientific novelty include a mathematical model of this system, which makes it possible to describe the process of its functioning using Markov chains. The difference from existing analogues is the unification of the description of the robot.

Keywords: control system, reinforcement learning, markov decision processes, neural networks, walking robot, artificial intelligence

The concept for designing a telecommunication subsystem of the traffic control system for passenger ships in Moscow

2024. T.12. № 1. id 1518

Shakhnov S.F. Smolentsev S.V. Butsanets A.A. Ivanova A.A.

DOI: 10.26102/2310-6018/2024.44.1.023

The increasing density of water transport traffic in Moscow requires regulating the movement of numerous tourist, pleasure and scheduled boats, as well as ensuring the necessary level of safety for passenger transportation. The government of Moscow is creating a system for organizing the movement of passenger ships (SOMPS), which is designed to solve these problems in the waters of the central pool of Moscow between locks No. 9 and No. 10. The SOMPS concept for Moscow based on the engineering-cybernetic approach is developed in the paper. The choice of a higher-level system (meta-system), the requirements of which the designed system must satisfy, is justified. A block diagram of the SOMPS, which includes five subsystems, is presented. The factors determining the properties of the designed system and the limitations affecting the functioning of the SOMPS are considered. The developed concept of the SOMPS makes it possible to isolate the optimal structure of the telecommunication subsystem, which provides the necessary channels of control, management and information exchange between the SOMPS elements and external systems, and to determine its tasks. The communication channels that are part of the telecommunication subsystem are described. External and internal factors that can affect the functioning of the designed SOMPS and its telecommunication subsystem are presented. The necessity of introducing a module for calculating the meeting point of vessels under bridges into the designed SOMPS is substantiated. Methods for transmitting alarm signals between the SOMPS and external systems are described. An example of the VHF repeaters placement along the passenger route No. 1 on the river bends is presented. A conclusion about the optimality of the telecommunication subsystem structure, which ensures the fulfillment of assigned tasks while minimizing the resources involved is made.

Keywords: telecommunication subsystem, system for organizing the movement of passenger ships, moscow water area, engineering-cybernetic approach, communication channel, traffic management center of Moscow Government, AIS, automated operator workstation

The use of optimization modeling in the management of the iterative process in team-oriented organizational system of the IT industry

2024. T.12. № 1. id 1517

Korchagin S.G.

DOI: 10.26102/2310-6018/2024.44.1.020

The article discusses the process of working on a project in a team-oriented IT system using optimization modeling of the iterative process of purposeful team activity based on management principles in an Agile-oriented organizational system. For such a system, the general principles of flexible development methodologies aimed at accelerating the creation of projects by dividing the final requirements into smaller parts in order to take into account feedback from stakeholders at each stage of work are transformed into problem-oriented ones, which allows us to highlight the management features from the point of view used to consider the process of working on the project. It is shown that these principles are initially implemented at the level of the management center, in particular, the project manager should be well-acquainted with them, then at the team level and only after that the developed optimization programs should be implemented. The results of their application are shown in the description of each management feature. For an objective assessment of the effectiveness of the developed software tools, a comparison of the results of work on similar projects is given. In one, only the principles of flexible project management methodologies were applied; in the other, developed software tools were additionally employed to manage the iterative process of goal achievement in a team-oriented organizational system.

Keywords: agile-oriented organizational system, team activity, multi-alternative optimization, expert assessment, management decision-making

Adaptive quantile regression

2024. T.12. № 1. id 1514

Tyurin A.S.

DOI: 10.26102/2310-6018/2024.44.1.016

The relevance of the research is due to the growing need for fast and accurate tools for building mathematical models. This paper discusses approaches to building adaptive quantile regression because selecting the optimal quantile during the training process can save a large amount of researcher's time. The correct choice of quantile can significantly improve the performance of the model on test datasets and, as a consequence, obtain more reliable predictions when such a mathematical model is actually used. The developed approach is a combination of modified quantile regression and gradient descent, which improves the adaptation of the model to different data. A detailed description of the developed algorithm is given. The paper also presents a comparison of the performance accuracy of the proposed model with traditional quantile regression and gradient descent along with their combinations. It also analyzes the training time of the models, including the number of training epochs. Experiments show that adaptive quantile regression exhibits improved accuracy with reduced training time. The results emphasize the effectiveness of this method in data analysis and prediction, opening new perspectives for more efficient and faster machine learning models.

Keywords: quantile regression, adaptive algorithm, gradient descent, mathematical modeling, numerical methods

Optimal management of combat

2024. T.12. № 1. id 1511

Belousova E.P.

DOI: 10.26102/2310-6018/2024.44.1.009

The article proposes a method for solving the problem of adapting a discrete inventory management model to the problem of combat operations of two armies. The aim is to identify the control effect on the linear system of difference equations, which allows it to be transferred from the initial to the final state in the specified parameters provided that costs are minimized. Discrete controlled processes play an important role in the theory and practice of optimal control since many planning tasks are described precisely by systems of difference equations. A system of equations of this type is characterized by a discrete type of control of the number of combat units at the current stage. Deliveries are formed at fixed intervals. The effectiveness of management is controlled (verified) by a quadratic quality criterion, which characterizes the cost of conducting combat operations. The criterion shows the total cost of supplies and maintenance of combat units, the change in the number of which is determined by three factors: the rate of losses as a result of hostilities, natural losses and the rate of receipt of reinforcements. The construction of an optimal control effect is carried out by the feedback method. It is noted that the solving this task is complicated by the fact that it is necessary to find among all possible solutions those that will make it possible to achieve your goals with the least expenditure of human and material resources. These costs are presented as functions of several variables, the values of which are known at the initial time. The article proves that in order to solve the problem of optimal resource management in relation to the case of combat operations of two armies, the feedback method is the most preferable. Several examples have been analyzed. The implementation of the feedback method clearly shows that a longer period of confrontation significantly reduces losses. The materials of the article are of practical value for strategic planning in the context of military conflicts.

Keywords: optimal control, discrete system, feedback principle, combat operations, control influence

Text authorship identification for open set of candidates in cybersecurity context

2024. T.12. № 1. id 1510

Romanov A.S.

DOI: 10.26102/2310-6018/2024.44.1.012

The paper considers the methods of authorship identification for fanfiction texts based on popular works of literature and cinema. The data for the study include texts from 5 popular topics of Ficbook online library. The most common is the closed set attribution task. Regarding practical issues, it can be assumed that the true author of an anonymous text will not always be included in the candidates set. Therefore, the process of author identification was regarded as a more complex version of the typical classification problem – the open set of authors. The proposed methods are based on the machine learning methods: fastText and One-Class SVM with informative features selection and statistical approaches of vector representation similarity measures. Statistical methods have proven to be the least effective even for the simple cross-thematic case. In comparison with the method based on One-Class SVM, the difference in accuracy reaches 15 %. For cross-thematic attribution, the average accuracy of the method based on the combination of One-Class SVM with feature selection and fastText was 85 %, while for the more complex task – classification within a group – it ranged from 75 to 78 % depending on the thematic group.

Keywords: text authorship attribution, fastText, machine learning, text analysis, information security

Distributed computing resource management method based on greedy strategy and efficient algorithms ontology

2024. T.12. № 1. id 1508

Klimenko A.B. Barinov A.A.

DOI: 10.26102/2310-6018/2024.44.1.018

Currently, managing computing resources in modern distributed computing systems is the relevant problem. As a result of infrastructure capability evolution, distributed computing can be organized in dynamic, heterogeneous and geographically distributed computing environments, examples of which are “fog” and “edge” ones. The dynamics of both load and topology imply the need to change the system configuration, namely, assigning user tasks to computing devices with the allocation of the necessary resources. The latter raises the issue of increasing the efficiency of the scheduler (broker), which facilitates management of network resources within the allocated fragment. Algorithmic and software schedulers are based on models and methods of scheduling theory and implement either simple heuristics, mathematical programming methods or metaheuristics. However, an analysis of publicly available problem statements has shown that, firstly, they are special cases and implement certain situations of computing resource distribution, and secondly, they do not fully reflect the properties of heterogeneity, geographical distribution and dynamics of computing environments. As part of this study, a general model of computing resource allocation problem is proposed with consideration to the listed properties, and a solution method using the subject ontology of metaheuristic methods is proposed. The feasibility of constructing and applying an ontology is shown using the example of analyzing the effectiveness of genetic algorithms depending on the values of the computing resource allocation problem parameters which is being solved.

Keywords: ontology, resource allocation, distributed computing, distributed computing management, resource management, optimization

Estimation of bacterioplankton abundance fluctuations in the vertical water column of Lake Baikal over a multi-year period

2024. T.12. № 1. id 1507

Burdukovskaya A.V. Belykh T.I. Rodionov A.V.

DOI: 10.26102/2310-6018/2024.44.1.013

The paper proposes two approaches to analyzing time series of bacterioplankton abundance in three different layers of the water column in Lake Baikal. In the first approach, the values of the seasonal component of the series are calculated using the moving average method, and additive and multiplicative models are constructed, from which the best models are selected on the basis of the calculated reliability coefficients. The seasonal component values in each of them are estimated. In the second one, correlation and regression analysis of joint changes in bacterioplankton abundance, temperature and lake water level is performed. Statistical hypotheses about the significance of correlation coefficients between the considered factors are put forward and tested. A mathematical model of multiple regression with inclusion of dummy variables describing the influence of seasonal fluctuations on changes in bacterioplankton abundance is constructed. Statistical assessment of the significance of the model and the factors included in the model is calculated. The results of correlation-regression analysis are interpreted in relation to the subject area under study. The findings can be used in predicting the amount of bacterioplankton in different periods of time, in making an ecological substantiation of the state of the lake, as well as in forecasting its microbiological state.

Keywords: time series, bacterioplankton, moving average method, seasonal component, correlation and regression analysis, multiple regression model, lake Baikal

Development of mathematical models and algorithms for optimizing the schedule of independent project activities

2024. T.12. № 1. id 1506

Rossikhina L.V.

DOI: 10.26102/2310-6018/2024.44.1.014

Planning is an important process for a project. The main planning processes include defining activities, planning resources, determining the duration of work, and developing a schedule. The paper examines projects with independent activities. The purpose of the study is to optimize project schedule by period. Three particular problems are considered. The first problem is to distribute activities over periods in order to achieve the maximum total effect of their implementation taking into account cost constraints in each period and the possibility of partial implementation of the activities in a given period. The solution algorithm is based on the Cost-Effect method. The validity of the proposed algorithm has been proved. The second problem deals with the distribution of work over periods with the prohibition of transferring part of the work to other periods and limitation of costs in each period. Based on the method of dichotomous programming, we propose an algorithm for solving the problem for two periods. For the number of periods greater than two, an approximate algorithm is suggested. For the case when information on unperformed activities in the course of project implementation changes, the problem of maximizing the total effect from the implementation of project activities in the current period is considered. Additionally, the effect from the implementation of a set of activities is visible after their completion and a certain effect manifests from the partial implementation of another set of activities. The effect obtained is proportional to the part of the amount of work performed. An algorithm for solving the problem based on obtaining parametric dependences of the total effect for each set of activities on the value of costs is proposed. The validity of the algorithm has been proved. Examples illustrating the application of the proposed algorithms are presented.

Keywords: project, work, period, effect, costs, resource, satchel problem, dichotomous programming method

Modeling and optimization of the placement of transmitting devices in a wireless communication system

2024. T.12. № 1. id 1504

Avetisyan T.V. Minaev K.A. Preobrazhenskiy A.P. Preobrazhenskiy Y.P.

DOI: 10.26102/2310-6018/2024.44.1.034

The paper considers the problem of signal propagation indoors. Several stages were considered in solving this problem. At the first stage, a model of electromagnetic wave propagation through the wall was built. An approach based on geometric optics was used. To calculate the degree of absorption, it is necessary to take into account the dielectric and magnetic permeability of the wall material. In order to automate the calculation process, a program was written in C++, which makes it possible to quickly determine the power values under given conditions. The attenuation of the radio signal depending on the angle of incidence on the wall is investigated. At the second stage, the tasks of determining the level of a propagating electromagnetic wave at various points inside the room are considered. At the third stage, the problem of optimizing the placement of the transmitting device inside the room is considered. A random search method was used with a sequential narrowing of the range of values. At the same time, the use of a local optimization method of the grid method was required. For each section of the grid, a local optimization method was used, which was the golden ratio method. As a result, after the implementation of several tens of thousands of iterations, the optimal placement of the transmitting device was determined. The scientific and practical significance of the work lies in the development of a complex algorithm for optimizing the placement of transmitting devices in the room based on a computational experiment.

Keywords: wireless communication, electromagnetic wave propagation, electromagnetic wave absorption, optimization, signal strength, signal attenuation

Modeling of a patch antenna in Comsol Multiphysics finite element analysis program

2024. T.12. № 1. id 1501

Cherkesov D.S. Kasatkina T.I.

DOI: 10.26102/2310-6018/2024.44.1.031

The article evaluates in detail the capabilities of patch antennas application based on the analysis of their advantages and disadvantages. The new patch antenna design was subjected to modeling, including description of its structure and creation of a three-dimensional model. The field distribution in the patch antenna geometry was obtained, which gives a complete picture of the influence of its structural elements on the electromagnetic properties. The directional diagram of the patch antenna is obtained, which reveals the angular features of its radiation. Plots of the gain of a single patch antenna, the gain of an 8×8 uniform array, and the gain of an 8×8 microstrip patch antenna plotted in dB-scale are constructed. It is shown that the design of the rectangular microstrip patch antenna with V-shaped notches provides better polarization at the edges compared to the center in the proposed patch antenna model, which can be a critical factor in real-world applications, especially in areas where communication quality is subject to external influences. The frequency at which this antenna resonates is 1,403 GHz, this allows for a wider bandwidth and improved impedance matching. These results emphasize the promising potential of the investigated patch antenna design in modern communication technologies and wireless data transmission systems.

Keywords: patch antenna, wireless communication, radiation pattern, gain factor, electromagnetic characteristics, antenna modeling, resonant frequency, size optimization, impedance matching

The model of optimal distribution of renewable resources in the management of the criminal threat prevention project and the methodology of its program realization

2024. T.12. № 1. id 1500

Zhirnov A.A. Ovchinskiy A.S. Makarov V.F. Gurlev I.V.

DOI: 10.26102/2310-6018/2024.44.1.010

The paper considers the applicability of a project approach to managing the activities of criminal investigation units. The problem of developing project management models for countering criminal threats is underscored. The disadvantages that do not make it possible to use the existing models of resource allocation in the subject area under study without significant adaptation are indicated. The criminal threat prevention project has been defined. The general parameters of such a project are given, which helps to determine the problem of renewable resource allocation. The approach to calculating the minimum amount of renewable resources required for the implementation of the k-th operation of the project. The problem of allocating renewable resources of the criminal threat prevention project, the purpose of which is to solve a crime, is defined. To solve this problem, an appropriate model of integer programming is proposed. A numerical example of the problem solution and a method of software implementation of the proposed model in a tabular processor are given. The features that the authors believe should be taken into account when applying the proposed model and methodology in practice and possible areas of future research are considered: the study of the project duration dependence type on the number of performers and other factors; the development of a knowledge base on the parameters of the conducted operations.

Keywords: criminal threat prevention project, allocation of renewable resources, project approach, critical path method, CPM, PERT

Forecasting and evaluation of energy generation at solar power plants: the state of the problem and development trends

2024. T.12. № 1. id 1499

Azhmukhamedov I.M. Loba I.S. Machueva D.A.

DOI: 10.26102/2310-6018/2024.44.1.008

The paper considers the relevant issues related to the problem of calculations and forecasting in the production of solar electricity as a renewable energy source. To detect problems, the initial data for modeling and their sources have been identified. Renewable energy sources are systematized and an example is given for each. An analysis of the state of the global energy market and the state of government policy in the field of energy in Russia has underscored the need to address solar energy issues and solve the problems of forecasting electricity generation. This is important not only due to the availability of resources, but also to environmental friendliness. The classification of existing models and methods for forecasting SES energy generation is examined. Existing methods allow calculations to predict the power generation capacity, but they give average figures for the year. New technological and innovative methods are required to solve the existing problem. The key factors and aspects of the introduction and operation of a solar power plant are presented. The main difficulty in forecasting is taking into account a variety of nonlinear characteristics. An attempt to solve this problem is proposed. An overview of the state of the problem and trends in the development of solar energy is made, among which the main problems are identified and solutions are outlined.

Keywords: solar energy, renewable energy sources, aspects and operation of a solar power plant implementation, forecasting solar energy generation, forecasting methods

Optimal control of an organizational and technical system taking into account the intensity of control actions application

2024. T.12. № 1. id 1497

Akhmedyanova G.F.

DOI: 10.26102/2310-6018/2024.44.1.019

Predictive management with all its errors and difficulties is still an effective means of providing an organizational and technical system with time to increase its readiness for changes in the situation. To formulate and solve the problem of optimal control of this process, the Fokker-Planck-Kolmogorov equation was used, which is the first approximation in the probabilistic description of random processes. To formulate the optimal control problem, the Letov criterion was modified, a coordinate-parametric approach was applied, and the obvious fact of an increase in management costs with a decrease in the time to improve the readiness of the organizational and technical system was taken into account in the form of the square of change rate in the probability density. The Euler-Ostrogradsky-Poisson equations are applied to the final Lagrangian. The resulting nonlinear equations were solved using the small parameter method. The study of the resulting solution proves that even with optimal control, the magnitude of control actions increases in proportion to the target value and duration of control (increasing the planning horizon), the increase occurs according to the cube of the exponential, that is, very slowly at the beginning of control and very sharply at the end, and a similar pattern of increase demonstrates the dependence of the control influences from the demand for management results, but it is expressed through hyperbolic functions.

Keywords: optimal control, fokker-Planck-Kolmogorov equation, probabilistic quality criteria, intensity of application of control actions, small parameter method

Integrated data storage system for geological laboratory experiments

2024. T.12. № 1. id 1495

Tishin N.R. Ozmidov O.R. Proletarsky A.V.

DOI: 10.26102/2310-6018/2024.44.1.007

The article examines the development of a new approach to storing and organizing the results of laboratory experiments with consideration to the features of their subsequent processing. To solve this problem, laboratory experiments are considered as structured data with unstructured parts. During the development of the system, the features of storing and processing laboratory test data were analyzed, after which the basic requirements for the system were formulated. The main data models were defined as well as the database entities. A standard relational data model has been chosen for storing structured data, and the storage of unstructured information such as experiment results or experiment parameters is implemented through the BJSON field. To solve the problem of providing secure access and creating an API for the system, the asynchronous FastAPI framework was chosen. The implementation of storing additional experiment files, which are located in the object storage and are associated with the experiment in the relational model through an additional entity, is also considered. The presented approach is notable for its flexibility to the structure of stored laboratory experiments, takes into account the features of geological laboratory experiments and also provides opportunities for complex meta-analysis of large volume of data. The system was tested and implemented into the technological process of the geotechnical laboratory at JSC MOSTDORGEOTREST.

Keywords: storage of geological laboratory experiment data, unstructured data, experiment results storage system, geoinformation system, database, geological environment, information resource, engineering geology

Acoustic emission diagnostics of hidden defects of multilayer printed circuit boards in electronic devices

2024. T.12. № 1. id 1493

Chernoverskaya V.V. Nguyen H.D. Lu N.T. The H.V. Uvaysov S.U.

DOI: 10.26102/2310-6018/2024.44.1.004

The article presents the results of the acoustic emission method application (AE) and machine learning algorithms in the problem of diagnosing defects in the stratification of a multilayer printed circuit board structure (MPB). A combination of physical and computational experiments is used to solve the problem. To conduct full-scale tests, the study uses a vibration stand to generate a load on the test object and receive acoustic emission signals. The computational experiment is carried out using mathematical modeling in a specialized ABAQUS environment. In order to obtain the best solution to the problem, an optimization problem is solved during the experiment to determine the frequency of the harmonic signal generated by the vibration stand with a view to receiving the maximum response of the MPB under review and unambiguous identification of the bundle defect. When conducting the numerical experiments, the effects and reactions (AE signals) of MPB were modeled at different frequencies of input vibration signals ranging from 100 to 2000 Hz. Full-scale experiments were conducted in the laboratory of control and testing of radioelectronic devices at the Department of KPRES of RTU MIREA. The results of the study have shown that the vibration frequency most effective for detecting a delamination defect equals 1500 Hz (a defect of almost rectangular shape with a size of 30×37 mm). Subsequently, this was confirmed by correlation analysis, which made it possible to identify the maximum differences between the acoustic emission signals of a suitable MPB sample and a sample with a delamination defect for the input vibration of a given frequency. The second part of the study deals with processing of the physical and computational experiment results, establishing the degree of adequacy of the obtained mathematical models to real samples of MPB and the processes occurring in them, as well as the use of machine learning algorithms for more reliable diagnosis of MPB defects. In the presented study, the random forest and the support vector machine learning (SVM) methods were employed as machine learning algorithms. Based on the results of their execution, the accuracy of the two algorithms was evaluated.

Keywords: acoustic emission, multilayer printed circuit board, hidden defects, structure stratification, modeling, physical experiment, machine learning algorithm, support vector machine method, random forest method, non-destructive testing

Feature selection methods for authorship attribution in cybersecurity context

2024. T.12. № 1. id 1489

Romanov A.S.

DOI: 10.26102/2310-6018/2024.44.1.001

This paper considers methods for authorship attribution of natural-language and artificially generated texts, which are important in the context of cybersecurity and intellectual property protection to prevent misinformation and fraud. The use of authorship methods is justified by the findings on the fastText and support vector method (SVM) effectiveness discussed in past studies. The feature selection algorithm is chosen based on the comparison of five different methods: genetic algorithm, forward and backward sequential methods, regularization selection and Shapley's method. The considered selection algorithms include heuristic methods, game theory elements and iterative algorithms. The regularisation-based algorithm is found to be the most efficient method, while methods based on complete brute-force selection are found to be inefficient for any set of authors. The regularization-based and SVM-based selection accuracy averaged 77 %, outperforming the other methods by between 3 and 10 % for an identical number of features. For the same tasks, the average accuracy of fastText is 84 %. A study was conducted to examine the robustness of the developed approach to generative samples. SVM proved to be more robust to model confounding. The maximum loss of accuracy for fastText was 16 % and for SVM was 12 %.

Keywords: feature selection, authorship attribution, machine learning, neural networks, text analysis, information security

Using artificial neural networks to perform segmentation of hip radiographs in the treatment of osteoarthritis

2024. T.12. № 1. id 1486

Akutin A.S. Goriakin M.V. Zubavlenko R.A. Pechenkin V.V. Solopekin D.A.

DOI: 10.26102/2310-6018/2024.44.1.011

Today, the X-ray analysis procedure makes it possible to detect osteoarthritis (OA) in the early stages of the disorder. The presence or absence of the disorder is detected only when it has already manifested, and X-ray diagnostics have been carried out. The use of automated procedures for analyzing X-ray images and the availability of archives of such information with a long history can improve the results of predicting complications in patients. The article describes the experience of developing an application for computer analysis of radiographs, which, based on deep learning methods, allows us to identify the risks of developing osteoarthritis of the hip joint. The archive of a specialized medical institute is used as a training sample. In order to increase the size of the training set of radiographs, a data augmentation method is used, which increases the variability of the original data and, in some cases, increases the recognition efficiency. The research uses a convolutional network (U-net) designed for image segmentation, which is trained on X-ray images of a specific medical institution. As part of a project to segment and analyze the geometric characteristics of X-ray images of the hip joints, the software to automate the recognition of the joint space size was developed, which helps to clarify the patient’s diagnosis and prognosis for the development of the pathology.

Keywords: convolutional neural network, image segmentation, machine learning, osteoarthritis, hip joint

Placing on-board equipment in the fuselage space of an unmanned aerial vehicle using a genetic algorithm

2024. T.12. № 1. id 1484

Gainutdinov R.R. Chermoshentsev S.F.

DOI: 10.26102/2310-6018/2024.44.1.021

The current stage of unmanned aircraft system development is characterized by the widespread introduction of automated and intelligent electronic systems. One of the most difficult and critical stages in the development of unmanned aerial vehicles is determining the optimal locations for placing on-board equipment in the fuselage space. To solve this problem, the approach for determining the optimal installation locations for on-board equipment in the fuselage space of an unmanned aerial vehicle is proposed. The approach is based on the use of a genetic algorithm. A meaningful and mathematical formulation of the problem of determining the optimal installation locations for on-board equipment in the fuselage space of an unmanned aerial vehicle is given. Criteria and restrictions have been developed. As optimization criteria, first of all, electromagnetic compatibility criteria are considered, which are characterized by minimizing the sensitivity of on-board equipment above the level of electromagnetic field strength at the installation sites of on-board equipment, as well as limiting the excess of the threshold level of susceptibility of on-board equipment over the electromagnetic environment resulting from electromagnetic influences or interactions. Additionally, criteria for minimizing the total weighted length of cable connections are considered, and the maximum load-carrying capacity of the fuselage compartments of an unmanned aerial vehicle is limited. The plan has been developed for the installation of on-board equipment in the fuselage space using a developed program that implements a genetic algorithm.

Keywords: placement, optimization, on-board equipment, genetic algorithm, unmanned aerial vehicle

Review of neural network models for solving the problems of predicting emergency situations and ensuring the safe operation of oil and gas wells

2024. T.12. № 1. id 1472

Sulavko A.E. Vasilyev V.I. Klinovenko S.A. Lozhnikov P.S. Suvyrin G.A. Guzairov M.B.

DOI: 10.26102/2310-6018/2024.44.1.017

An analytical study was carried out on the problem of preventing emergency situations and predictive diagnostics of equipment during hydrocarbon production in oil and gas fields as well as the ways to solve this problem by means of artificial intelligence based on deep neural networks. One of the key factors hindering the development of predictive equipment diagnostic systems is the lack of data describing pre-emergency situations, which is necessary for high-quality training of neural network models. An analysis of recent publications and research on the subject of telemetry data analysis and emergency recognition is provided. Neural network models are considered that can be used to predict the failure of pumping and compressor equipment and other units. Cases of the use of neural network models specially trained to solve this problem, as well as neural network models used in other tasks but analyzing similar data structures, were studied. The issue of transfer learning is raised to adapt neural network models originally developed and trained for other areas to use in the area under consideration in order to reduce the sample size when training industrial artificial intelligence. A comparison of the achieved results was carried out, and the advantages and disadvantages of existing technical solutions were identified.

Keywords: artificial neural networks, predictive diagnostics, machine learning, time series, telemetry, maintenance, data sets

Dysarthria speech recognition by phonemes using hidden Markov models

2024. T.12. № 1. id 1471

Bredikhin B.A. Antor M. Khlebnikov N.A. Melnikov A.V. Bachurin M.V.

DOI: 10.26102/2310-6018/2024.44.1.002

The relevance of the paper is due to the difficulties of oral interaction between people with speech disorders and normotypic interlocutors as well as the low quality of abnormal speech recognition by standard speech recognition systems and the inability to create a system capable of processing any speech disorders. In this regard, this article is aimed at developing a method for automatic recognition of dysarthric speech using a pre-trained neural network for recognizing phonemes and hidden Markov models for converting phonemes into text and subsequent correction of recognition results using a search in the space of acceptable words of the nearest Levenshtein word and a dynamic algorithm for splitting the output of the model into separate words. The main advantage of using hidden Markov models in comparison with neural networks is the small size of the training data set collected individually for each user, as well as the ease of training the model further in case of progressive speech disorders. The data set for model training is described, and recommendations for collecting and marking data for model training are given. The effectiveness of the proposed method is tested on an individual data set recorded by a person with dysarthria; the recognition quality is compared with neural network models trained on the data set used. The materials of the article are of practical value for creating an augmented communication system for people with speech disorders.

Keywords: hidden Markov models, dysarthria, automatic speech recognition, phonemes recognition, phoneme correction

Development of hybrid atmospheric-underwater optical communication system

2024. T.12. № 1. id 1468

Ali M. Saklakov V.M.

DOI: 10.26102/2310-6018/2024.44.1.032

Underwater optical wireless communications are promising and future-oriented wireless carriers to support underwater activities focused on 5G and beyond (5GB) wireless systems. The main challenges for the deployment of underwater applications are the physicochemical properties and strong turbulence in the transmission channel. Therefore, this paper analyzes the end-to-end performance of a hybrid free space optics (FSO) and underwater wireless visible light communication (UVLC) system under intensity modulation or direct detection (IM/DD) in a method considering a pulse amplitude modulation (PAM) scheme. In this study, a fading model with Gamma-Gamma (GG) distribution is used to deal with channel conditions with moderate and strong turbulence, and the links are designed by combining plane wave modeling in the corresponding links, respectively. The proposed performance methods excel in higher achievable data rates with minimal delay response and improves network connectivity in real-time monitoring scenarios compared to conventional underwater wireless communication techniques. The simulation results provide reliable estimates of system performance metrics such as average bit error rate (ABER) and probability of failure (Pout) in the presence of pointing errors. Finally, this paper uses a Monte Carlo approach for best curve fitting and validate the numerical expression with simulation results.

Keywords: 5G and 5GB networks, cooperative communication, optical communication, underwater communication, underwater sensor networks (USNs), VLC light communication

Functional near-infrared spectroscopy (fNIRS) as a hybrid system: a review

2024. T.12. № 1. id 1459

Samandari A.M.

DOI: 10.26102/2310-6018/2024.44.1.005

Sensor devices and biomedical imaging technologies used in clinical application scenarios are essential for providing a comprehensive portrait of patients’ state, but these technologies, despite their outstanding advantages, have their inherent disadvantages. Beginning with the principle of complementary images of medical imaging techniques, this review examines the functional near- infrared spectroscopy (fNIRS) technique and its use as a hybrid system. The fNIRS technology delivers impressive results in terms of the biological signal classification accuracy, but its use as a hybrid system with electroencephalography (EEG) and electromyography (EMG) achieved better results because it has become a complementary tool to fill the deficit of the common technology with it, and this has been highlighted in this review. The results show that the superiority in the biological signal classification accuracy provided by hybrid systems from fNIRS with EEG and EMG would provide a comprehensive and objective assessment of the patients’ state from the stage of illness to healing. In conclusion, we have no indication from the scientific studies of the previous four years (2020–2023) that demonstrate which of the hybrid systems is better than others when used in clinical practice, and this encourages further in-depth studies to validate the combination of methods to prove their success and preference.

Keywords: HBCIs, fNIRS, fMRI, EEG, EMG, MEG

Finite element modeling of thermohydraulic processes by the porous body method

2024. T.12. № 1. id 1457

Yaurov S.V. Danilov A.D. Gusev K.Y.

DOI: 10.26102/2310-6018/2024.44.1.006

The paper considers the best-known models of a porous body used to simplify the performance of thermohydraulic calculations by the finite element method. The main approaches and dependencies when using the porous body model in calculations are shown. The results of thermohydraulic calculations using the Darcy porous body model are presented. The calculation of a heat exchanger with spirally wound tubes was performed, the calculation of a complex technological system consisting of mechanical filters of different configurations was performed. The discrepancies between the calculated and actual parameters of the equipment are determined. The use of a porous body model as a hydraulic analogue of equipment using the example of mechanical filters and a heat exchanger showed acceptable results (deviations from the design values range from 0,1 % to 10 %). These discrepancies are related to the accuracy/correctness of the selection of porous body resistance laws (dependencies). The use of the porous body approach in modeling the operating modes of technological systems including equipment with a complex design is explained, first of all, when it is required to predict the operating modes of the system as a whole from the result of computational modeling, but local processes occurring inside the equipment are not. Secondly, when it is necessary to reduce the time for performing calculations with low available power capabilities of computers. However, the proposed approach has disadvantages, in particular, the procedure for determining the degree of porosity of the simulated object and the laws of hydraulic resistance selected from empirical dependencies is quite complex.

Keywords: porous body model, complex technological systems, heat exchanger, finite element method, hydraulic resistance, mechanical filters

Automated generation of a complete data model of production facilities in a unified information environment

2024. T.12. № 1. id 1456

Filimonova A.A. Chizhov M.I. Vetokhin V.V. Sobenina O.V.

DOI: 10.26102/2310-6018/2024.44.1.003

The article examines the problem of developing an integration platform to facilitate end-to-end business processes supporting the life cycle of heterogeneous information objects. The platform topology is chosen according to the functionality of the integrated systems and the structure of the information object. To create a unified enterprise information environment, various topologies are considered, including peer-to-peer, message broker, centralized, and hybrid topologies. The basis for the description of an object is a complete data model, including defining attributes and transformation rules corresponding to each of the integrated systems. Using the object model of the information support system for digital products and special templates, a methodology for forming policies, methods and documents (PMD) and organizing a unified digital environment of the enterprise is proposed. However, to solve this problem, the development of a specialized integration platform is required which is capable of processing data from production facilities on a centralized basis and facilitating their interaction in a unified information environment. Such a platform must take into account the characteristics of each system component and ensure the security of information exchange; it also should be able to scale and adapt to the changing needs of the enterprise. In addition, this article discusses in detail various topologies for creating a unified enterprise information space. Peer-to-peer, message brokered, centralized, and hybrid topologies are included. Each of these topologies has its own characteristics and advantages, and the choice of the optimal one depends on the requirements and characteristics of a particular enterprise. To successfully implement integration and create a unified digital environment of the enterprise, it is suggested to use an object model of an information support system for digital products. This model helps to structure information and determine the relationships between various components of the system. Furthermore, the article proposes a methodology for the formation of PMD, which is the basis for organizing a unified digital environment of the enterprise. This methodology takes into account the requirements for security, consistency and efficiency of the system and also ensures standardization and consistency of processes within the enterprise.

Keywords: information production facilities, integration, digital environment, full data model, process automation