Models, systems, networks in economics, technology, nature and society
Current Issue
No 4 (2025)
MODELS, SYSTEMS, NETWORKS IN ECONOMICS AND MANAGEMENT
RESEARCH OF INTERACTION MECHANISMS OF AGRICULTURAL EQUIPMENT MARKET PARTICIPANTS IN THE CONDITIONS OF IMPORT SUBSTITUTION
Abstract
Background. In the current complex geopolitical situation, with the imposition of sanctions against Russia and the withdrawal of leading suppliers of sophisticated equipment from the market, it is important to strengthen Russia's technological sovereignty through import substitution, particularly in the agricultural sector, as agricultural products are crucial for the country's food security. This situation necessitates the development of new and sustainable mechanisms for cooperation between agricultural machinery manufacturers, farmers, the government, and scientific institutions. The problem is all the more relevant because since the end of the last century, the Russian agricultural machinery market has become highly dependent on imports due to the closure of domestic production facilities. Therefore, the purpose of this article is to develop and systematize mechanisms for effective cooperation between participants in the agricultural machinery market in the context of import substitution. To achieve this goal, it is necessary to identify and classify the main market participants; analyze existing models and forms of interaction; identify market problems; and propose mechanisms for interaction. Materials and methods. The methodological basis of the study was based on a review and analysis of the works of domestic and foreign researchers in the field of the functioning and development of the agricultural sector of the Russian economy in general and the market for agricultural machinery sales in particular. The study used a systematic and logical approach, and the following research methods were employed: the method of scientific abstraction, the monographic method, the logical method, and the method of expert assessments. The study examined the legislative and regulatory acts of the Russian Federation, as well as the data from the official websites of the legislative and executive authorities in the field of scientific, technological, and innovative development of the country's agricultural sector. Results. The current state of the agricultural machinery market in Russia has been analyzed, and key participants and their roles have been identified. The mechanisms of interaction between them have been described and analyzed. Key problems in interaction have been identified. Organizational and economic mechanisms of interaction are proposed. Conclusions. Based on the analysis of the interaction mechanisms between market participants, the main problems of the market have been identified, and solutions to these problems have been proposed, such as the development of a cluster model and the creation of consortia.
5-17
PROSPECTS FOR THE DEVELOPMENT OF E-COMMERCE IN RUSSIA:CHALLENGES AND GROWTH DRIVERS
Abstract
Background. The relevance of the study is determined by the need for a comprehensive analysis of the factors influencing the development of the market, as well as the assessment of the risks and prospects for sustainable growth in the industry. The purpose of the study is to identify the key drivers and challenges of e-commerce development in Russia, determine promising areas for its growth, and assess the impact of technological, institutional, and socioeconomic factors on the formation of the industry's competitive advantages in the medium term (2025–2030). Materials and methods. The methodological framework of the research is based on a combination of general scientific and specialized approaches to the analysis of socio-economic processes. The research employed methods of systematization, comparison, logical and analytical generalization, as well as statistical, content, and trend analysis. Secondary marketing research was used. The information base of the research includes materials from leading analytical and consulting centers (INFOLine, IBS Real Estate, Data Insight, SCG, T-Bank eCommerce, GfK Russia, SBS Consulting), official statistical data, reports from online stores and marketplaces, expert publications, academic studies on the digital economy and e-commerce, as well as the author’s own analytical observations. Results. As a result of the conducted research, the structure and dynamics of the development of the e-commerce market in Russia were analyzed, key trends, directions and problems of its functioning were identified. Key areas of transformation have been identified: digitalization of business processes, consolidation of the market around leading marketplaces, development of niche and regional platforms, as well as strengthening the importance of the ESG agenda and business social responsibility. It is determined that the further growth of the sector will depend on the level of technological maturity of companies, their innovation activity, the quality of customer experience, the effectiveness of regulation and the ability of market participants to adapt to the changing conditions of the digital economy. Conclusions. The Russian e-commerce market is facing a number of systemic challenges: high concentration of capital, rising costs, staff shortages, regulatory risks, and environmental constraints. These factors require companies to increase their technological adaptability, improve their business models, and integrate the principles of sustainable development. Successfully overcoming these challenges will ensure the formation of competitive and sustainable ecosystems that can become drivers of economic growth in the digital age.
18-43
DEVELOPMENT OF THE PASSENGER RAILWAYTRANSPORTATION SYSTEM IN THE REPUBLIC OF CRIMEA
Abstract
Background. Given the special importance that passenger rail transport has for the economy of the Republic of Crimea, it is advisable to conduct research on the population's demand for rail travel to the Republic of Crimea, both retrospectively for the period of 2022–2024 and prospectively until 2030, taking into account the commissioning of a new railway section along the coast of the Sea of Azov. To determine the role and place of railway transport in the economy of the Republic of Crimea, which since 2022 has become the only mode of transport that provides mass transportation of passengers and significant volumes of cargo to the peninsula. To show the relationship between railway transport and the Crimean Peninsula in the areas. Materials and methods. The work used general methods of scientific knowledge, methods of comparative analysis and statistical methods to identify the projected demand of the population for rail travel to the Republic of Crimea for the period up to 2030, taking into account the commissioning of a new railway section. Results. The study showed the important role and perspective of passenger transportation development for the region of the Republic of Crimea. The role of suburban passenger rail services for the Republic of Crimea is also significant. The calculation of passenger traffic attraction coefficients between the constituent entities of the Russian Federation and the Republic of Crimea showed that the regions with the greatest potential for increasing the population's demand for railway travel to the Republic of Crimea. Conclusions. According to the author's forecast, the population's demand for railway travel to the Republic of Crimea in 2030 will amount to 4,223.8 thousand people, an increase of 2.1 times compared to 2024. The author proposes the organization of new railway routes to the Republic of Crimea. The author considers these routes to be promising and of practical significance for the constituent entities of the Russian Federation, especially the Republic of Crimea and the new regions of the LPR, DPR, Zaporizhzhia, and Kherson. Further research is being conducted on the possible future volume of passenger traffic on the proposed routes. This research can be used by regional authorities, state corporations, and business representatives to develop and implement infrastructure projects in the transportation sector and related economic sectors.
44-67
IMPROVING THE QUALITY OF REGIONAL TYPOLOGIZATION IN THE RUSSIAN FEDERATION BASED ON INNOVATION DEVELOPMENT INDICATORS USING NEURAL NETWORKS
Abstract
Background. As regional socio-economic data become increasingly complex and the role of innovation grows, accurate regional typologization gains importance as a tool for analysis and for substantiating differentiated regional economic policy. However, the application of classical clustering methods to the analysis of regional innovation development is constrained by high data dimensionality and the sensitivity of results to initialization. The aim of this study is to develop and test a methodological approach to improving the quality and stability of clustering of the regions of the Russian Federation based on innovation development indicators by combining an autoencoder neural network with the K-means algorithm. Materials and methods. The empirical basis of the study consists of official statistical data on the innovation development of the regions of the Russian Federation, represented by a set of 30 indicators. A two-stage approach is employed, including preliminary nonlinear dimensionality reduction using an autoencoder followed by clustering using the K-means method. The quality and stability of clustering are assessed using the silhouette index and the Davies – Bouldin index across a series of 10 experiments. Results. The results demonstrate that the use of an autoencoder provides a statistically stable improvement in clustering quality according to both evaluation metrics compared to the classical K-means algorithm and reduces the impact of initialization on regional grouping outcomes. Conclusions. The proposed approach enhances the methodological robustness of regional typologization based on multidimensional innovation development indicators and establishes a foundation for more reliable economic interpretation of cluster structures and for supporting decision-making in regional and sectoral economics.
68-83
MODEL OF REGIONAL INNOVATIVE INVESTMENT HEALTHCARE ECOSYSTEM BASED ON THE MEDICAL TECHNOLOGICAL UNIVERSITY
Abstract
Background. The paper discusses the urgent problem of building an effective innovation and investment ecosystem for a list of projects for the development and launch of complex medical equipment MedTech on the market. Materials and methods. Based on the formal model of innovation and investment activities for the implementation of a portfolio of projects, a new optimization problem of interdisciplinary sifting is set, the solution of which consists in consolidating resources for maximum coverage of projects by implementing activities of minimal cost, taking into account time constraints. Results. A solution to the problem of interdisciplinary sifting in a network interpretation is proposed. For a typical life cycle of a project for the development of high-tech medical equipment, types of activities and organizations capable of organizing logistical support for products at different levels of technological readiness are identified. Conclusions. To build an effective innovation and investment ecosystem of projects for the development of high-tech medical equipment, the most suitable organization is a medical technology university with the possibility of serial production. This should be a subordinate institution of the Ministry of Health with its own clinical base and an extensive variable innovation infrastructure.
84-95
MODELS, SYSTEMS, MECHANISMS IN THE TECHNIQUE
MODELING OF KOVAZHNY FLOW AND TAYLOR – GREEN VORTEX ON PHYSICS-INFORMED RADIAL BASIS FUNCTION NETWORKS
Abstract
Background. An analysis of physics-informed neural networks for solving partial differential equations has been conducted, and the advantages of physics-informed radial basis function networks have been demonstrated. The practical application of physics-informed neural networks requires the development of fast learning algorithms for such networks and the expansion of the classes of solvable problems. The aims of this work is to develop new algorithms for training physics-informed radial basis function networks for modeling the Kovazhny flow and Taylor – Green vortex and to create extensions to the TensorFlow library for implementing networks. Materials and methods. A new Sophia algorithm (Second-order Clipped Stochastic Optimization) has been adapted for training physics-informed radial basis function networks. The adaptation consists in training not only the network weights but also the parameters of radial basis functions and using a simplified approximation of the Hessian diagonal. For the first time, algorithms have been developed for modeling the Kovazhny flow and Taylor – Green vortex on physics-informed radial basis function networks. Extensions to the TensorFlow library have been developed that implement radial basis function network layers, loss functions, and an adapted training algorithm. Results. The results of modeling the Kovazhny flow and Taylor – Green vortex showed that when modeling the Taylor – Green vortex, the adapted Sophia algorithm reduced the number of iterations by 3.2 times compared to the popular ADAM method, and when modeling the Kovazhny flow, by 3.7 times. At the same time, the adapted Sophia algorithm slightly increases the training time compared to the fastest Levenberg – Marquardt algorithm, but is much easier to implement. Conclusions. The areas of further research are the development of new PIRBN learning algorithms and the solution of new model problems.
96-113
THE TECHNIQUE OF FORMING A TRAINING SET AND APPLYING A NEURAL NETWORK DETECTOR TO RECOGNITION OF GRAPHICAL USER INTERFACE ELEMENTS
Abstract
Background. One of the points of increasing the degree of automation of the software development process is testing, in which there is still a significant proportion of manual labor, especially in testing user interfaces. The article proposes a method for solving the problem of recognizing the basic elements of graphical user interfaces, which can be used as one of the stages of automated software testing. An overview of the existing marked-up datasets is provided with an assessment of their suitability for training algorithms designed for automated software testing systems. Materials and methods. The research suggests a technique for applying a neural network detector to the task of recognizing graphical interface elements, including the formation of a set of training and test data. A new data set has been formed containing about two thousand images of user interface elements, divided into 5 classes, and a solution is proposed in the form of a neural network detector based on the YOLOv5m model. Results. According to the training results, the quality metrics of the model were: mAP50-95 – 0.6, accuracy – 0.8, completeness – 0.8. Conclusion. The developed solution has increased recognition accuracy, which makes it possible to use it in automated testing systems for applications with a user interface.
114-124
PROCESSING TRANSPORT INFRASTRUCTURE MONITORING DATA FOR DIGITAL TRAFFIC MANAGEMENT PROJECTS
Abstract
Background. This study aims to develop a new approach to processing transport infrastructure monitoring data for the creation of digital traffic management projects. The aim of the study is to develop an effective methodology for data analysis and interpretation, ensuring greater efficiency of decision-making in the field of traffic management. Materials and methods. The study is based on the use of the "Trassa" road laboratory, equipped with equipment for a comprehensive survey of roads. The developed "AnchorLab" software is used, which implements the proposed methodology for processing road information and localizing zones requiring modification of the traffic management scheme. Results. The proposed approach provides a comprehensive analysis of the geometric characteristics of roads, the existing traffic management scheme, regulatory requirements, and the performance characteristics of the transport system to develop design options that reduce the risk of road accidents and improve the overall capacity of the transport network. Conclusions. The use of the proposed approach improves the level of information support for decisionmaking in the field of traffic management. The results of the work can be used by designers and road services during the survey, design, and maintenance of transport infrastructure.
125-138
NUMERICAL STUDY OF THE EFFECT OF TEMPERATURE CONDITIONS AND HEAT TREATMENT TIME OF RUBBER PRODUCTS OF VARIOUS THICKNESSES ON THE VULCANIZATION PROCESS
Abstract
Background. Rubber products are widely used in various industries, but their production is often accompanied by challenges due to the low thermal conductivity of rubber compounds. This leads to energy wastage and unnecessary costs during the experimental determination of operating parameters. To address this issue, approaches to mathematical modeling of the vulcanization process are being actively developed. The aim of this study is to investigate the effect of temperature field and holding time of rubber compound in mold on the state of cure, in order to make design and technological decisions for the production of rubber products more efficiently. Materials and methods. The study focuses on the vulcanization processes of cork and membranes at various operating conditions, using numerical modeling of temperature fields and state of cure through ANSYS finite element analysis. Results. Graphs of changes in the state of cure and temperature of rubber products during the vulcanization process are obtained. Graphical dependencies of the standard deviation of the state of cure in relation to the volume of parts at different times after extraction of products from the mold are also obtained. The results indicate an uneven course of the reaction in thick-walled products. It has been found that for the specific rubber compound, the unevenness of the temperature field has little effect on the final state of cure. The standard deviation of the state of cure by volume of the finished product did not exceed 2 %. However, the holding time of the rubber mixture in the mold significantly affects vulcanization, and should be individually selected for each product. Therefore, suboptimal extraction conditions can lead to a standard deviation in vulcanization of over 20 %. Conclusions. The results of the presented numerical modeling make it possible to evaluate the kinetics of the vulcanization process for making design and technological decisions in the production of rubber products.
139-152
CONTINUOUS-TIME MARKOV MODELSOF CYBER ATTACKS: IMITATION MODELING
Abstract
Background. The article is devoted to the development of a theoretical framework for cybersecurity mechanisms and methods used in the creation and maintenance of modern information systems. The aim of the work is to estimate the average time to security failure, a key indicator reflecting the time characteristics of a system's resilience to cyberattacks. Materials and methods. Among the diverse approaches used in this field, stochastic simulation plays a key role, standing out as one of the most effective analysis tools. The basis of the cyberattack models we study are Markov and semi-Markov chains, in which the dynamics of state transitions of the system is described using nonstationary Poisson and uniform distributions. Results. The results obtained demonstrate that the unsteadiness of attack flows significantly affects this indicator, which underlines the importance of taking into account temporal variability when assessing the security of real information systems. Conclusions. The use of stochastic modeling makes it possible to describe the dynamics of "attack – reflection" type processes.
153-164
POTENTIAL REDUCTION IN THE DIMENSIONALITY OF THE PRICE FORECASTING TASK ON THE STOCK EXCHANGE BY SPLITTING THE DATA INTO INTERVALS WITH DIFFERENT REGRESSION SIGNS
Abstract
Background. The aim of the work is to evaluate the reduction in the dimensionality of the problem of forecasting price fluctuations in the market in relation to the completely unpredictable chaos of white noise. Materials and methods. The white noise is simulated by a software random number generator. The Shannon entropy, estimated in the Hamming distance space, is used as a dimension metric. Results. It is shown that the simplest procedures that link the Hurst exponent to the regression sign can reduce the entropy of the model's code states by more than 5 times. Conclusions. The assessment of the entropy of market conditions, combined with the use of the empirical Hurst indicator, should help identify persistent and anti-persistent market conditions, based on a background of about one month in depth, and allow for more reliable forecasts.
165-174
SUPPORT FOR MANAGEMENT DECISION-MAKING IN SPATIALLY DISTRIBUTED ORGANIZATIONAL SYSTEMS BASED ON A RISK-BASED APPROACH
Abstract
Background. This paper examines the problem of improving the management efficiency of geographically distributed organizational systems characterized by complex spatial relationships and transitive risk effects. Traditional risk management approaches poorly account for spatial heterogeneity and do not provide interpretable support for management decisions. Materials and methods. To build an adaptive risk-based management system, a spatially oriented graph model is proposed. It is based on the principles of system approach and risk management, scenario-cognitive modeling, and risk mapping practices. The model is formalized as a directed weighted graph, where nodes (risks, KPIs, and control actions) have multidimensional attributes, including fuzzy and spatiotemporal damage assessments. Results. A method for cascading risk impact assessment on performance indicators and an algorithm for synthesizing control actions taking into account territorial localization are developed. The possibility of integrating the model with regional geoportals for spatial analysis, identifying vulnerable zones, and substantiating priority measures is demonstrated. Conclusions. The proposed approach provides interpretable support for management decisions under conditions of uncertainty and spatial heterogeneity, expands the focus of risk management from loss minimization to opportunity management, and is compatible with modern GIS technologies for scenario analysis and strategic planning.
175-188
DEVELOPMENT OF AN INFORMATION AND MEASUREMENT SYSTEM FOR ASSESSING THE DYNAMICS OF CHRONIC APICAL PERIODONTITIS
Abstract
Background. The aim of the work is to develop a joulemetric informationmeasuring system for the rapid assessment of the dynamics of the inflammatory process in chronic apical periodontitis and to establish an objective criterion for determining the possibility of permanent obturation of the root canals. Materials and methods. The main research tool was the information-measuring system developed by the authors. The primary parameter to be evaluated was the current work index, which characterizes the activity of electrochemical processes in the periapical tissues. To assess the statistical significance of the parameter change dynamics, nonparametric criteria for related samples were used. Results. The structure of an information and measuring system has been developed, including a specialized two-electrode sensor with an isolated active electrode based on an endodontic instrument, a current pulse generation unit with an amplitude of 0.1 μA and a duration of 3 s, an interelectrode voltage measurement unit and data processing software. A prospective experimental study was conducted with the participation of 14 patients with a verified diagnosis of chronic apical periodontitis. Four series of measurements of the current indicator were performed for each patient with an interval of seven days. For statistical analysis, the nonparametric Friedman criterion was used for related samples, followed by Wilcoxon analysis with the Bonferonni correction. In the course of the study, the authors established a statistically significant decrease in the current indicator during treatment. An objective criterion for the completion of the active phase of inflammation has been established – a decrease in the value of the current indicator by 75–80 % of the initial value, which is a marker for the appointment of permanent obturation of the root canal. Conclusions. Thus, the relationship between the current indicator and the activity of the inflammatory process has been confirmed, and the developed system can be used for dynamic assessment of the course of chronic apical periodontitis.
189-198