РУС
ENG
№ 3(38)'2025
Aims and Scope
Founder and Publisher
Editorial board
Licensing terms
Privacy Statement
Plagiarism policy
Publication ethics
Archiving Policy
Subscription
Instructions for authors
Reviewing proccess
Copyright Notice
License agreement
Article processing charges
All issues
Search
Contacts
Section
ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING
Хао Х
именованных сущностей BERT -BiLSTM-CRF для автоматического распознава ния и извлече -
UDC: 519.4
DOI: 10.24412/2413-7383-2025-3-38-4-10
Language: Russian
Abstract: В ответ на быстрый рост объема литературы в области тугоплавких высокоэнтропийных сплавов (RHEAs) и низкую эффективность извлечения ключевой информации в данной статье предлагается полуавтоматический рабочий процесс извлечения информации. Метод использует большую
Keywords: Refractory high-entropy alloys, named entity recognition, large language model, BERT-BiLSTM-CRF.
D. I. Dzeboev
TENSORIAL ALGEBRA
UDC: 512.554.1
DOI: 10.24412/2413-7383-2025-3-38-11-18
Language: Russian
Abstract: The aim of this study is to develop and substantiate a new universal method for defining multiplication, applicable to a wide class of algebraic systems. The proposed solution is based on the introduction of Tensorial Algebra, where the product is determined through a third -order algebra tensor. This approach makes it possible to specify or to learn multiplication rules in arbitrary algebras and ensures the preservation of dimensionality in operations. It is shown that the proposed method generalizes hypercomplex systems and opens up prospec ts for applications in the theory of non - associative algebras, as well as in problems of mathematical modeling and computational algebra.
Keywords: tensorial algebra, hypercomplex numbers, non-associative algebras, multiplication structures, algebraic systems
С. М. Селякова, А. В. Ниценко, Н. К. Базиян-Кухто
INFORMATION SUPPORT FOR THE DECISION-MAKING PROCESS IN THE TREATMENT OF PATIENTS WITH CRITICAL LOWER LIMB ISCHEMIA
UDC: 004.89+616.1
DOI: 10.24412/2413-7383-2025-3-38-19-33
Language: Russian
Abstract: This article presents the development of the "AngioExpert" decision support system (DSS), aimed at improving the efficacy of treatment for patients with critical limb ischemia (CLI). The system facilitates optimization of surgical management strategies and mitigates risks associated with postoperative complications and mortality. The paper details the system's core functionality, overall architecture, functional model, and key user interface components. It further examines the specific clinical decision-making tasks addressed by the system's algorithms.
Keywords: clinical decision support, decision support system (DSS), vascular surgery, critical limb ischemia (CLI).
D. N. Tumakov1, A. R. Berdigulov1, D. I. Khairullina1, I. I. Bariev2
APPLICATION OF LENET-5-TYPE NEURAL NETWORK ENSEMBLES FOR DIAGNOSING PNEUMONIA IN LUNG X- RAY IMAGES
UDC: 004.032.26, 004.932.2
DOI: 10.24412/2413-7383- 2025-3-38-34-50
Language: Russian
Abstract: Ensembles of shallow neural networks for identifying pneumonia from X-ray images are considered. The ensemble's resulting response is constructed according to the principle: if at least one of its neural networks identifies pneumonia, the ensemble also ret urns the answer "pneumonia." This approach to aggregating the results of individual neural networks maximizes the recall of the "pneumonia" class. To minimize the deterioration in precision that is inevitable with this aggregation approach, the ensemble's neural networks are all of the same type —LeNet-5. Six neural network architectures are proposed, each receiving 256 by 256 pixel images as input and having different convolutional matrix sizes. The architectures of the networks considered also have differe nt depths. On the test set, the recall for the optimal ensembles was 0.991, while the precision varied between 0.934 and 0.936.
Keywords: neural network ensembles, LeNet-5, recall, precision, pneumonia.
Section
MATHEMATICAL MODELING, NUMERICAL METHODS AND SOFTWARE SYSTEMS
V. N. Belovodskiy, S. L. Bukin
ON THE MOVEMENTS OF A MATERIAL PARTICLE ON A HORIZONTAL TRANSPORTING MACHINE OF A NONLINEAR TYPE PERFORMING SUB- AND SUPERHARMONIC VIBRATIONS
UDC: 621; 534.014.4
DOI: 10.24412/2413-7383-2025-3-38-52-64
Language: Russian
Abstract: A single-mass horizontal transporting machine with a bilinear characteristic of elastic elements and inertial harmonic excitation is considered and its vibrations in the zones of sub - and superharmonic resonances are analyzed. The possibility of excitation in these frequency ranges of oscillations close to optimal biharmonic has been established. Using the example of a mathematical model of motions of a material particle along a horizontal vibrating surface, a comparative analysis of the velocity of its movement is performed. It is demonstrated the possibility to provide a particle velocity close to the maximum by implementing a subharmonic resonance of the order of 1:2.
Keywords: vibrating transporting machine, sub- and superharmonic resonances, material particle
V. S. Solod, V. M. Zuyev, S. B. Ivanova
EVALUATION OF THE POSSIBILITIES OF USING NEURAL NETWORK MODELS TO OPTIMIZING REINFORCING STEEL COOLING TECHNOLOGY
UDC: 621.771.07:004.94
DOI: 10.24412/2413-7383-2025-3-38-65-77
Language: Russian
Abstract: The aim of the work is to evaluate the possibilities of using simple artificial neural networks (ANN) for modeling complex metallurgical p roduction processes. The application of ANN for describing the technology of accelerated cooling (AC) of reinforcing bars in the rolling mill flow, designed to form a specified level of mechanical properties of finished rolled products, is considered. Based on the literature analysis, a set of input and target parameters, the ANN architecture are determined. A data set for training the ANN, obtained by calculating according to the mathematical model of the authors, is developed. The training process was carried out with a variable number of neurons and target parameters. The analysis of the results of model evaluation is carried out. The direction of further research is proposed.
Keywords: reinforcement profile, accelerated cooling, perceptron, network model
S. A. Fedorov, Y. A. Bezus, A. E. Rybalka
POLARITON EXCITATIONS IN A TOPOLOGICALLY ORDERED IMPERFECT HEXAGONAL ARRAY OF MICROPORES WITH A PRIMITIVE LATTICE
UDC: 539.2+535
DOI: 10.24412/2413-7383-2025-3-38-78-86
Language: Russian
Abstract: In this paper, a microscopic examination of polariton excitations in a non -ideal topologically ordered nanolattice of micropores with hexagonal symmetry is carried out. The specifics of the energy spectrum of the quasiparticles under consideration are studied, due to both the configuration dependence of the parameters included in the Hamiltonian system and its symmetry.
Keywords: hexagonal symmetry, polariton structures, Bose-Einstein condensation.
Section
SYSTEM ANALYSIS, CONTROL AND INFORMATION PROCESSING, STATISTICS
2JSC M.A. Kartsev Research Institute of Computing Systems
MIVAR KNOWLEDGE BASE AUTOMATIC GENERATION METHOD FOR SOLVING MANAGEMENT TASKS IN STATE SPACE
UDC: 004.82+007.52
DOI: 10.24412/2413-7383-2025-3-38-88-99
Language: Russian
Abstract: The paper is devoted to the development of a method for the automatic generation of mivar knowledge bases for solving management tasks in the State Space. The method is used to process information in order to prepare a decision -making model (mivar knowledge base) for systems with Artificial Intelligence. T he method offers an approach to preparing source data and knowledge for generation and a model of an automatic knowledge base generator based on the Metagraph approach to describing hybrid intelligent information systems.
Keywords: mivar approach, metagraph approach, mivar knowledge base, mivar network, metagraph, expert system, intelligent control system, action planning, state space
2JSC M. A. Kartsev Research Institute of Computing Systems
DECISION-MAKING MODEL FOR MULTITASKING WAREHOUSE ROBOTS BASED ON A FOUR-LAYER COLLABORATIVE OPTIMIZATION FRAMEWORK
UDC: 004.89 + 007.52 + 681.518
DOI: 10.24412/2413-7383-2025-3-38-100-112
Language: Russian
Abstract: This paper proposes an original model based on a four -layer collaborative optimization framework, which integrates the mivar decision -making system and multi -level optimization mechanisms, overcoming the trade -off between dynamic response and computational complexity in traditional methods. A dynamic priority mechanism is employed to differentiate between urgent and large-scale cargo; a dual-mode optimization engine is considered, which addresses the issue of combinatorial explosion in limited -scale tasks ( n≤6) through global enumeration and a hybrid strategy combining nearest neighbor and 2-opt approaches. Experiments conducted in a simulated warehouse environment with 100 heterogeneous orders demonstrated that the model effectively generates 42 task group s (single/two/three/four-task groups accounting for 16.7%/40.5%/26.2%/16.6% respectively), with a total transport distance of 5436 meters (averaging 129.4 meters per group). This structure provides an optimization paradigm for intelligent warehousing systems, combining rules and models, and future research will be exten ded to scenarios involving heterogeneous robot planning and multi -robot transport of oversized cargo.
Keywords: mivar, mivar technology, dynamic task allocation, mivar decision-making system, logical AI, intelligent warehousing, multi-task robots, four-layer optimization
A. V. Nicenko, V. Ju. Shelepov, S. A. Bolshakova
AUTOMATIC TEXT SEGMENTATION INTO SEMANTICALLY HOMOGENEOUS FRAGMENTS (PARAGRAPHS)
UDC: 004.89
DOI: 10.24412/2413-7383-2025-3-38-113-123
Language: Russian
Abstract: The work is devoted to the topical topic of segmentation of the Russian -language text into semantically homog eneous fragments. The paper provides an overview of a number of existing approaches to this task. An algorithm is proposed for automatically dividing text into paragraphs as thematically homogeneous fragments by using a ratio that takes into account the fr equency of occurrence of a word and the length of the text segment where it occurs. Using this software, experiments were conducted and the proposed algorithm was compared with other text segmentation methods. As a result, it was found that the proposed al gorithm demonstrates better performance than other compared approaches.
Keywords: semantic text segmentation, paragraph, word frequency, segment of occurrence.
Section
AUTOMATION AND CONTROL OF TECHNOLOGICAL PROCESSES AND PRODUCTION
Technical University named after Academician M.D. Millionshchikov"
OPTIMIZATION OF ENERGY EFFICIENCY AND ENVIRONMENTAL FRIENDLINESS OF CEMENT PRODUCTION BASED ON THE UPGRADED AUTOMATIC CONTROL SYSTEM OF THE BALL DRUM MILL
UDC: 681.51
DOI: 10.24412/2413-7383-2025-3-38-125-137
Language: Russian
Abstract: The article is devoted to reducing the environmental impact of cement production through the introduction of clinker -free additives (slags, ash) and an upgraded ball mill speed CAP. A mathematical model has been developed that includes the dynamics of an electric drive with a PID controller, thermal protection of windings and bearings, as well as algorithms for processing sensor data. Simintech and laboratory experiments have confirmed the stability of the system (overshoot <5%, transition time — 2.5 s) and the effectiveness of overheating protection. The solution reduces CO₂ emissions by 15-20% by optimizing grinding and using recycled materials.
Keywords: clinkerless cement, ball mill, automation, PID controller, thermal protection, mathematical modeling
Technical University named after Academician M.D. Millionshchikov"
AUTOMATED ENERGY SAVING SYSTEMS AS KEY TECHNOLOGIES OF THE ELECTRIC POWER INDUSTRY
UDC: 519.8:658.5128
DOI: 10.24412/2413-7383-2025-3-38-138-147
Language: Russian
Abstract: The article discusses modern automated energy saving systems (ASE), their architecture, key technologies and practical aspects of implementation. The analysis of efficiency, economic benefits and real cases of implementation in various industries is presented. Special attention is paid to promising areas of development, including the use of artificial intelligence and the Internet of Things.
Keywords: automated systems, energy saving, artificial intelligence, digital twin, energy efficiency.
А. А. Личман, О. Ю. Чередникова
ского метода ARIMA и алгоритма машинного обучения случайный лес был внедрен алгоритм
UDC: 004.032.26
DOI: 10.24412/2413-7383- 2025-3-38-148-156
Language: Russian
Abstract: The introduction of automated control systems has affected almost all key sectors of the economy. From the end of the last century to the present day, the environment in which production facilities exist is constantly changing, and automated control systems are forced to change with it. For example, in the last ten years, partial intellectualization of production facilities has made it possible to optimize costs. Predictive analytics is one of the most effective optimization tools. In this article, an intellectualization algorithm was implemented using the statistical ARIMA method and the random forest machine learning algorithm. The results of the algorithm implementation are analyzed.
Keywords: ARIMA, automated control system, intellectualization, time series, optimization.
Section
МЕТОДЫ И СИСТЕМЫ ЗАЩИТЫ ИНФОРМАЦИИ, ИНФОРМАЦИОННАЯ БЕЗОПАСНОСТЬ
A. A. Levashov, D. A. Garkusha
VISUALIZATION OF PRODUCTION AND TECHNOLOGICAL PROCESSES OF THE SEA PORT
UDC: 004.921
DOI: 10.24412/2413-7383-2025-3-38-158-171
Language: Russian
Abstract: The article proposes requirements focused on the tasks of optimizing resources and speeding up decision- making at the operational, tactical and strategic levels of seaport management. The requirements for displaying statistical and dynamic information, interactivity, performance, integration, and user experience (UX) are formulated. A comparative analysis of web technologies for interactive 2D graphics is carried out. The choice of the React, Redux and Konva technology stack is justified.js to implement the established requirements. The principles of implementing key functions, including drag-and-drop for manual object placement planning, are presented. The results of the work confirm the effectiveness of the proposed method and the selected tools for creating scalable, productive and interactive web visualization systems to improve the manageability and logistical efficiency of seaports.
Keywords: web visualization, seaport, production and technological processes, interactivity, drag-and-drop, React.js, Redux, Konva.js, logistics optimization.
I. A. Tretiakov, M. V. Babicheva, Ia. I. Rushechnikov, V. V. Danilov
ELEMENTS OF AUTOMATED SCIENTIFIC RESEARCH SYSTEMS FOR RADIO-ELECTRONIC ENVIRONMENT
UDC: 004.056:681.5:621.37
DOI: 10.24412/2413-7383-2025-3-38-172-181
Language: Russian
Abstract: A hardware and software module has been developed to detect unauthorized electromagnetic radiation in the range of 85 -115 MHz, an electrical circuit has been finalized, programs have been written for firmware of the ATtiny85 microcontroller in the algorithmic C language and for docking the module with a personal computer in Pascal, test tests and calibration of the developed hardware and software module have been performed on radio microphone layouts, a comparative analysis of the results has been carried out, obtained using the developed module with the results obtained, using the SDR Sharp software and hardware module. The approach of implementing horizontal scaling of components of automated radio monitoring systems makes it possib le to improve the quality characteristics of the radio system without high costs, as well as expand the list of tasks solved with its help.
Keywords: TEMPEST, technical channels of information leakage, information security, ASRS, radio environment, radio monitoring, software-defined radio systems
U. A. Kasparov, V. M. Zuev
EXERCITATIO MODELLORUM VISIONIS COMPUTATRICIS AD DETECTIONEM ET SEGMENTATIONEM CELLULARUM STEMICARUMВ
UDC: 004.93
DOI: 10.24412/2413-7383-2025-3-38-183-189
Language: Russian
Abstract: This article examines key approaches to training computer vision models for stem cell detection and segmentation. A methodology for working with high-resolution images for image segmentation tasks is developed and tested.
Keywords: location: detection, image, segmentation, high resolution Hic articulus modos clavis ad exempla visionis computatralis ad detectionem et segmentationem cellularum primordialium exercenda examinat. Methodologia ad imagines altae resolutionis ad
1Federal State Budgetary Institution "V.K. Gusak Institute of Emergency and Reconstructive
INTEGRATION OF MULTIPHYSICS MODELING AND MACHINE LEARNING FOR THE DEVELOPMENT OF MAGNETIC CELL DELIVERY SYSTEMS
UDC: 004.032.26+004.8+537.622+616-089
DOI: 10.24412/2413-7383-2025-3-38-190-197
Language: Russian
Abstract: оптимизация. Magnetic-guided delivery of mesenchymal stem cells (MSCs) is a promising approach for deep tissue therapy, such as for femoral neck fractures, but it requires powerful external magnetic systems. This work presents an approach for their automated design using machine learning. Based on a multiphysics model, a target magnetic force factor of B⋅∇B≥26.5 Т2/м was established. Using Bayesian optimization integrated with finite element modeling, a system configuration capable of generating a force factor of 21.35 Т2/м at a target depth of 5 cm was found. The study demonstrates that the synergy of machine learning and modeling is a powerful tool for automating the development of complex medical devices.
Keywords: mesenchymal stem cells, magnetic cell delivery, regenerative medicine, machine learning, multiphysics modeling, finite element modeling, magnetic fields, biomedical devices, optimization.
