metadata of articles for the last 2 years
Работая с нашим сайтом, вы даете свое согласие на использование файлов cookie. Это необходимо для нормального функционирования сайта, показа целевой рекламы и анализа трафика. Статистика использования сайта отправляется в «Яндекс» и «Google»
Научный журнал Моделирование, оптимизация и информационные технологииThe scientific journal Modeling, Optimization and Information Technology
Online media
issn 2310-6018

metadata of articles for the last 2 years

Synthesis of neural network architecture for ship pattern recognition based on pre-training technology

2022. T.10. № 2. id 1148
Gulamov A.A.   Konarev D.I.  

DOI: 10.26102/2310-6018/2022.37.2.011

The relevance of the article is due to the information and communication support of navigation by monitoring river vessels using video surveillance cameras. The main goal is to recognize ships in images, for which the application of neural networks has potential. The aim of the paper is to study the performance indicators of vessel recognition by means of available pre-trained networks after their additional training for the assigned tasks and to select the most efficient network. The research considers various pre-trained neural networks. The input data for the networks are ship images. The training sample was collected manually and includes two independent DataSets with images of river vessels and many other objects apart from ships. The networks were built and further trained with the aid of Keras and TensorFlow machine learning libraries. The employment of pre-trained convolutional artificial neural networks for pattern recognition problems and the advantages of utilizing such networks over synthesizing a neural network from scratch are presented. The architecture of efficient pre-trained VGG16 neural network is described in detail. An experiment was conducted in additional training of available pre-trained convolutional neural networks for the assigned task. The efficiency of various pre-trained neural networks was evaluated in terms of the percentage of correct pattern recognition cases on the test set. The most efficient neural network for ship pattern recognition tasks has been selected. NASNetMobile and NASNetLarge networks have shown the maximum accuracy. However, the minimum image size that these networks can work with is larger than for other available networks and the great number of parameters in the convolutional layers of these networks causes a significant increase in retraining and operation time than for other available networks. Concurrently, VGG16 neural network with a small number of parameters and a short time for additional training has proven to be highly efficient which is why it is recommended for the purposes of ship pattern recognition.

Keywords: artificial neural networks, pre-trained networks, convolutional neural networks, keras, tensorFlow, google Colaboratory, VGG16, NASNetMobile, NASNetLarge

Algorithmization of repeated query optimization in cloud databases with the aid of computer training

2022. T.10. № 1. id 1147
Almusawi O.   Kravets O.J.  

DOI: 10.26102/2310-6018/2022.36.1.020

In cloud environments, hardware configuration, data usage, and workload distribution are constantly changing. These changes make it difficult for the query optimizer of the cloud database management system to choose the optimal query execution plan (QEP). In scientific literature, it was proposed to re-optimize the query during its execution for the purpose of optimizing it with a more accurate cost estimate. However, some of these optimizations cannot provide performance gains in terms of query response time or monetary costs, which are the two optimization goals for cloud databases, and may have a negative impact on performance due to overhead. This raises the question of how to determine when the optimization is efficient. The aim of the study is to develop a method of repeated query optimization that uses computer training. The key idea of the algorithm is to employ past query executions to learn how to predict the effectiveness of query re-optimization, and this is done in order to help the query optimizer avoid unnecessary re-optimization of queries for future ones. The method runs the query step-by-step, utilizing a computer training model, to predict whether re-optimization of the query will be useful after the stage is completed, and calls the query optimizer to automatically perform re-optimization. An experimental evaluation of the effectiveness is to be carried out.

Keywords: repeated query optimization, cloud databases, computer training, multi-stage query, automation of execution

Network-centric motion control algorithm for a group of mobile robots

2022. T.10. № 1. id 1146
Diane S.   Iskhakov A.Y.   Iskhakova A.O.  

DOI: 10.26102/2310-6018/2022.36.1.026

When solving reconnaissance and tactical tasks, methods of mobile robots group application show high efficiency. Robotic groups, based on the network-centric system of control, are characterized by the superiority of intelligence systems over enemy intelligence systems, including the reliability, timeliness and accuracy of the extracted information. At the same time, planning the group trajectory that facilitates communications maintenance with the aim of transmitting control signals becomes a priority in the implementation of such secure systems. The paper proposes a scientific approach to handling the task of mobile robots motion planning under the conditions of providing mechanisms for secure interaction between the agents of a robotic system, using steganographic methods to hide control signals. Previously, the authors developed and tested methods and algorithms as well as software solutions for concealing control signals and the facts of their transmission within the process of intellectual interaction of a robotic systems group when they address a common problem, as well as verification of agents by a dynamic tuple of identification attributes. In the ongoing study, we put forward the software for trajectory planning of a heterogeneous multi-agent robotic system on the condition of maintaining communications to perform the transfer of control signals.

Keywords: information security, control systems in robotic devices, communications security, multi-agent robotic system, network-centric control

Locally one-dimensional method for the transfer equation of a continuous medium with distributed parameters on a network-like domain

2022. T.10. № 2. id 1141
Tran D.  

DOI: 10.26102/2310-6018/2022.37.2.008

The paper considers a wide range of issues related to the solution of an initial-boundary value problem for a parabolic partial differential equation with a multidimensional space variable belonging to the Euclidean space and changing on a network-like domain. The mathematical model describing the process of transferring a continuous medium over a network carrier is determined by the formalism of the initial-boundary value problem. An idea that has become classical is further developed for the case when a network-like region is a directed bounded graph, i.e., a collection of a finite number of segments connected to each other by means of end points. The study employs classical approximations of evolutionary differential equations of the 2-nd order as well as non-classical approximations of differential relations illustrated by generalized Kirchhoff conditions at the branching points of a network-like region (nodal points of the region). When using difference approximations of the initial-boundary value problem operator, the approximation error and stability conditions for the difference scheme are established. The characteristic properties of the locally one-dimensional method and the sweep method utilized to solve the stated problem are studied. An algorithm for the numerical solution of the stated problem is proposed, a computer program is designed, and a computational experiment is carried out on a series of applied problems. The findings are of interest in the analysis of applied problems of multiphase continuum media transfer along network-like 3D carriers.

Keywords: initial-boundary value transfer problem, network (directed graph), continuous medium transfer, difference scheme, locally one-dimensional method

Component optimization of an evolving digital management environment in organizational systems

2022. T.10. № 2. id 1140
Ryndin N.A.  

DOI: 10.26102/2310-6018/2022.37.2.013

The article deals with the formation of optimization models for component optimization of a digital management environment in organizational systems based on probabilistic assessments of component functional requirements fulfilment and estimation of the components implementation influence parameters on the achievement of established requirements. Methods and algorithms for calculating the parameters of digital environment components implementation influence on the achievement of established requirements are considered. As a principal method, it is proposed to use multi-alternative optimization and choosing the option of component integration into a single digital environment that provides the specified requirements for the functioning of a digitalized organizational system. Special attention is given to evaluating the functionality of digital environment components to determine the suitability of a component or a need to replace it in case of non-compliance with the specified requirements at the stage of its development. Boundary conditions for the transition from the stage of functioning to the stage of digital environment development are regarded in terms of fulfilling the established requirements for organizational system parameters and developing control actions: further operation of the component or its replacement; introduction of a new component to meet new system requirements; adjustment of component functionality under the conditions of the system unchanged structure.

Keywords: digital environment, probabilistic estimates, life cycle, resource allocation, component optimization

Safe distance determining algorithm in terms of the thrust force dependence on the aircraft speed during takeoff-run

2022. T.10. № 2. id 1139
Nguyen V.   Uvaysov S.U.   Florova I.A.   Rychkova O.V.  

DOI: 10.26102/2310-6018/2022.37.2.001

The importance of air communication between various points of the globe in the modern world is difficult to overestimate. Yet the employment of this means of transport is associated with high risks for passengers, crew, cargo and the aircraft itself due to the possibility of serious accidents at all phases of the flight, but especially during takeoff and landing. This article presents a physical and mathematical model of an aircraft takeoff-run. Its analysis helps to avoid accidents in the event of emergency situations. This model enables the creation of an electronic device for monitoring takeoff dynamic characteristics and warning the aircraft crew about arising inconsistencies. The article presents differential equations describing the dynamic characteristics of the aircraft during takeoff-run. Additionally, solutions of these equations are obtained, which explicitly determine the functional dependencies of the distance necessary for a safe takeoff on the time elapsed since the start of the takeoff-run. The influence of external factors, such as ambient air temperature, wind speed during takeoff and runway slope on the calculated characteristics is considered. As an example, the article also offers the results of an emergency takeoff modeling with the aid of modern software (flight simulator flightgear 2020.3, GeoGebra mathematical program). From the authors’ point of view, the materials of the article may be of practical value for developers of non-embedded on-board control devices, as well as for users of these devices.

Keywords: takeoff, takeoff-run, runway, gravity, friction force, lifting force, normal reaction force of the support, thrust force, drag force, satellite receiver

Analysis of the COVID-19 pandemic impact on the development of human capital in the region using machine learning algorithms

2022. T.10. № 1. id 1137
Kashirina I.L.   Azarnova T.V.   Bondarenko Y.V.  

DOI: 10.26102/2310-6018/2022.36.1.004

The COVID-19 pandemic has had a major impact on the formation and development of human capital through its negative effect on education and public health. This disease has already claimed hundreds of thousands of lives, caused long-term health problems and deprived many of them of access to quality education. In this regard, during the COVID-19 pandemic, it is of great importance to design modern and accurate methods for analyzing, modeling and predicting the dynamics of the spread of this disease, which enable to identify factors that significantly affect the spread of the infection. The article discusses the stages of constructing machine learning models for conducting a predicative analysis of the COVID-19 incidence, which makes it possible to study the dynamics of the spread of this virus at the regional level, identify the influence of various factors on the severity, the duration of the disease, and subsequently create timely scenarios for managing the human capital of the region in order to reduce the negative impact of the pandemic. To devise the methods, a large array of depersonalized data on the spread of COVID-19 in the Voronezh region, provided by the Voronezh Regional Clinical Consultative and Diagnostic Center, was used. The article presents the results of an exploratory analysis of the available data, highlights additional features that can be employed to build machine learning models and develops methods for interactive visualization and forecasting of COVID-19 dynamics.

Keywords: human capital, COVID-19, machine learning, trend forecasting, exploratory data analysis

Means of optical and radio frequency identification in the technological process of automated control of mobile on-board equipment circulation

2022. T.10. № 1. id 1136
Styskin M.M.   Stepanov P.V.   Zheltov S.Y.   Sokolov B.V.   Ronzhin A.L.  

DOI: 10.26102/2310-6018/2022.36.1.003

The relevance of developing multicomponent means of identifying mobile on-board equipment is due to the increasing requirements for the quality and routine maintenance speed of aircraft. This article analyzes the problems of existing methods for controlling mobile on-board equipment circulation of civil aviation aircraft. The tasks, solved by the equipment circulation control system in the course of loading and unloading mobile on-board kitchen equipment, its transportation, downtime in remote warehouses at destination airports, are formulated. The main problem, solved by the mobile on-board kitchen equipment circulation control system, is the identification of each equipment unit in a reliable manner. To address this issue, various options for automatic contactless identification of mobile equipment, based on RFID, QR-coding, Bluetooth technologies, are proposed. As a result, each unit of equipment receives its own unique digital identifier, which is stored and tracked throughout the entire life cycle in the enterprise information system. The architecture of the mobile equipment circulation control system is presented, including subsystems for generating visual tags, programming identification tags (RFID, BLE), registration, disposal, control of loading, unloading, stocktaking, search, and etc. The application of the mobile equipment circulation control system, which is a part of the technological process of aircraft service maintenance at an airport, is outlined. The advantages of the devised system for controlling the loading/unloading of on-board equipment are demonstrated.

Keywords: circulation control system, mobile equipment, identification tags, visual tags, QR code, RFID, bluetooth, BLE

The algorithm for pulse measurement using a human’s and fetus’s phonocardiogram without classifying heart sounds

2022. T.10. № 1. id 1135
Kosteley Y.V.   Zhdanov D.S.   Borovskoy I.G.  

DOI: 10.26102/2310-6018/2022.36.1.018

The relevance of the research is due to the prospects of phonocardiogram employment for daily monitoring of a fetus’s and mother’s state by designing inexpensive portable devices. Currently, there are no software and hardware systems that completely solve this problem. In this regard, the article presents the algorithm for pulse measurement by means of a phonocardiogram for signal processing in the streaming mode. The algorithm feature is a human’s and fetus’s pulse measurement based on the heart sound segmentation without classifying them into the first and second heart sounds, as well as detecting them under the conditions of other physiological and mechanical sounds, absence of one of the heart sounds in the signal, and deviations in the segmented sound boundaries from the actual localization of the cardiovascular system. For assessing the credibility of the results, the pulse values, obtained by the developed algorithm and reference methods, were compared. The pulse values, calculated by two experts with the aid of a phonocardiogram, were used as a standard for comparison. The average relative deviation between the findings and the reference method results does not exceed 3 %. The article materials are of practical value for the design of a fetus and human state daily monitoring systems.

Keywords: phonocardiogram, cardiointervalogram, heart sound segmentation, heart rate, signal processing

Adaptive control of magnetotherapy by means of biotechnical feedback on the impedance of biologically active points in physiotherapy of ischemic patients

2022. T.10. № 1. id 1134
Miroshnikov A.V.   Petrunina E.V.   Pavlenko A.V.   Protasova Z.O.   Shekhine M.T.   Shulga L.V.  

DOI: 10.26102/2310-6018/2022.36.1.023

The article examines a biotechnical system of rehabilitation and treatment of ischemic patients. A generalized structural diagram of rehabilitation of patients with high ischemic risk by exposing them to magnetic fields with controlled biotropic parameters is presented. The scheme allows building a model of the living system functional state and implementing adaptive control of magnetotherapy through biotechnical feedback on surrogate markers. A biotechnical system of magnetotherapy for patients with coronary heart disease has been developed. Biotechnical feedback was introduced into the system, which enabled it to adapt the magnetotherapy program to the functional state of the patient and correct it during the therapeutic session. Adjusting the therapeutic magnetic field parameters made it possible to increase the therapeutic effect of the physiotherapeutic procedure, reduce the adaptation and negative reactions of the body to magnetotherapy, and plan magnetotherapy programs. To implement the feedback, which ensures the adaptation of the magnetic field biotropic parameters to the functional state of the patient, we used information about the impedance of biologically active points and ischemic risk classifiers, the descriptors of which were determined on the basis of this information. An algorithm for controlling the biotropic parameters of the magnetic field by means of the functional state multimodal classifiers of the patient and a fuzzy inference module, designed to correct the biotropic parameters of the magnetic field in the course of a magnetic therapy session, is given. In a clinical setting, it was shown that the application of adaptive magnetic therapy is an effective method of treating patients with angina pectoris II and III of functional classes (85% and 77%, respectively), which is 14% and 15% higher than the corresponding results in the control group.

Keywords: ischemic risk, adaptive magnetotherapy, biotechnical system, biologically active point, neural network, fuzzy control module, base of fuzzy decision rules, algorithm

Optimization of a supply chain digital thread in the practice of managing organizational systems

2022. T.10. № 1. id 1132
Mukha V.V.  

DOI: 10.26102/2310-6018/2022.36.1.027

The article discusses an approach based on the use of methods for modeling, optimization and expert assessment in the development of a supply chain digital thread in the practice of managing organizational systems. It is shown that a digital thread needs to ensure, on the one hand, the movement of material flows within the supply chain and, on the other hand, the exchange of information flows during the interaction of the control center digital platforms and objects of the organizational system. The task is to make a managerial decision regarding the necessity to modernize the operating system of the logistics process digital control in comparison with the results of multivariate modeling and optimization of the digital thread structure. First of all, a model assessment of information exchange efficiency indicators is carried out in the case of basic structures of interaction between digital platforms of the control center and objects that form the supply chain: centralized and decentralized. In order to expand the expert evaluation possibilities of the new version of the logistics process digital management, simulation of the material flows movement processes and data exchange is undertaken to optimize the three-level cluster structure of the digital thread. All these stages of making a managerial decision are combined into a structural scheme for choosing an effective option for implementing a digital thread of supply chains.

Keywords: organizational system, management, logistics process, digital thread, optimization, simulation, expert assessment

Study of current chronograms, structure of dissociation, recombination and equilibrium regions in problems of non-stationary 1:1 electrolyte transfer in membrane systems using a mathematical model

2022. T.10. № 1. id 1131
Kovalenko A.V.   Gudza V.A.   Chubyr N.O.   Khromykh A.A.   Urtenov M.K.  

DOI: 10.26102/2310-6018/2022.36.1.028

The joint research of the dissociation-recombination reaction and space charge along with their effect on the transfer of 1:1 electrolyte ions appears to be a relevant issue. The article is a theoretical study of the dissociation, recombination and equilibrium areas and the features of salt ion transfer in each of these areas using the method of mathematical modeling. In the article, for the first time, on the basis of a mathematical model of non-stationary transfer of 1:1 electrolyte, the main regularities of the influence of the dissociation and recombination non-catalytic reaction on the transfer of 1:1 salt ions and electroconvection are theoretically established. In particular, the chronograms of the current density with and without taking into account the dissociation/recombination reaction of water, the structure of the dissociation regions, recombination and equilibrium were examined, the dependences on the input parameters were determined: the initial concentration, the potential sweep rate. It has been shown that in the boundary layers of ion-exchange membranes, the dissociation reaction prevails over the recombination reaction due to the fact that in these regions the electric field strength takes such high values that the electric field breaks the water molecules and separates the Н+ and ОН- ions, preventing them from recombining. It has been demonstrated for the first time that in the middle part of the desalination channel, a region is formed where the recombination of H+ and OH- ions predominates. This reaction is local in nature, so all H+ and OH- ions cannot recombine at the same time. As a result, in the region of recombination, an excess of H+ ions emerges on the one side and OH- on the other side, in other words, an electric double layer is developed in the middle part of the desalination channel, and the recombination region is rather narrow. The obtained theoretical results and conclusions can be applied to analyze the operation of electrodialysis machine desalination channels.

Keywords: electrolyte, membrane, dissociation, recombination, electroconvection, desalination channel

Implementation of numerical methods for assessing the probabilistic-temporal characteristics of a set of operations in the form of a problem-oriented software package

2022. T.10. № 2. id 1130
Oleinikova S.A.   Selishchev I.A.   Nedikova T.N.  

DOI: 10.26102/2310-6018/2022.37.2.012

The object of research in the article is a project represented by a set of sequential-parallel operations with a random duration. A feature of the project class under study is the dependence of the execution time of any operation on the executive assigned to it. The goal is to assess the probabilistic and temporal characteristics of the project, which include, first of all, the duration of service. Due to the stochastic nature of operation duration, this characteristic will be random. Based on this, it is required to estimate the mathematical expectation of the investigated random variable in the first instance. In addition, the variance of the random variable, its distribution law, and the ability to estimate the probability of project completion in a given timeframe (i.e., the probability of the investigated random variable falling within a given timeframe) pose an interest. To achieve this goal, the article presents a software package consisting of an application for assessing the parameters of individual operations, a simulation system for assessing the duration of the entire project and a database that stores information about the required entities. All components of the software package are implemented using the Java language; MySql was chosen as the DBMS. As a result, it became possible to evaluate the parameters of individual operations, transfer them to the simulation environment for conducting an experiment to assess the necessary characteristics of the duration of the entire project and save the obtained values in the database for their possible further use.

Keywords: probabilistic-temporal characteristics, sequential-parallel operations, numerical method, software package, simulation modeling

Information support in scientific paper preparation management on the basis of intelligent review analysis

2022. T.10. № 1. id 1128
Latypova V.A.  

DOI: 10.26102/2310-6018/2022.36.1.005

Intelligent technologies are actively entering different fields of activity. These technologies are also beginning to be utilized at certain stages of scientific paper preparation management process to organize information support for peer reviewers and editors of the academic journals for the purpose of reducing their labour costs and accelerating the procedure of paper publication. However, tools, implementing intelligent technologies, are mostly used in major publishing houses due to the fact that their application requires strong computing resource consumption in their work. This entails the need for appropriate hardware and large financial expenditure. For that reason, the employment of intelligent technologies is nowadays limited for most of other academic journals. To deal with this problem, an approach to decision-making information support in scientific paper preparation management on the basis of intelligent review analysis is suggested in this article. The testing of the proposed measure to information support at the Aeterna Research and Publication Centre has proven its effectiveness by the example of paper preparation management in Academicheskaya publitsistika (Academic Publicism) and Innovatsionnaya nauka (Innovative Science) Journals.

Keywords: information support, scientific paper, paper preparation management, peer review, intelligent review analysis

Simulation modeling features of capacitive sensor measurement circuits

2022. T.10. № 1. id 1127
Mishukov S.V.  

DOI: 10.26102/2310-6018/2022.36.1.017

The main universal and effective method of researching numerous technical systems and objects is the implementation of simulation modeling in special software, for example, SimInTech (Simulation In Technic), which is a unique software package that makes it possible to combine calculation projects and schemes with various mathematical, physical, electrical, hydraulic and other processes in real time. This feature of the SimInTech environment enables to simulate measuring circuits based on the original method of determining the capacitive sensor parameters, which consists in a two-stage measurement algorithm in steady-state and transient modes of the circuit under study. Using the example of the specified measuring circuit, the features of constructing its model by means of a built-in programming language, the development of custom blocks of a capacitive sensor, an operational amplifier, a sampling and storage device are shown. In addition, a study of the random component of the sensor parameter measure error, when the accuracy of voltage sampling changes, was carried out and the possibility of performing measurements with an error not exceeding 0.5% was confirmed. The materials of the article are of practical value for specialists in the field of control and measuring technology since the application of the SimInTech dynamic modeling environment for building models of measuring circuits will facilitate theoretical research relating to multi-element circuits and objects represented by them.

Keywords: capacitive sensor, measuring circuit, simulation modeling, multi-element circuit, submodel

A training device for the analysis of anomaly detection methods based on machine learning theory

2022. T.10. № 1. id 1122
Grekov M.M.  

DOI: 10.26102/2310-6018/2022.36.1.019

Nowadays, the timely detection of new malicious attacks on computer networks appears to be a relevant issue. In this regard, it is necessary to develop anomaly detection methods that enable the identification of unknown attacks. The paper presents a model of a training device for analyzing anomaly detection methods in reliance on machine learning theory. A model has been developed for generating datasets with characteristics of real network traffic by means of a generative adversarial neural network. The generated dataset can be employed to train and test detection models while the sample emulates the features of a real network, which increases the efficiency of anomaly detection. The training device can also use publicly available datasets: NSL-KDD, CICIDS2017. Support vector machine, k-nearest neighbors, naive Bayes, logistic regression, decision trees, random forest, k-means are utilized as training methods, and a multilayer neural network, based on the PyTorch library, is implemented. The training device simplifies the process of analyzing machine learning methods, applied to obtain anomaly detection models. The developed software product facilitates not only training and testing with the aid of publicly available datasets, but also provides the ability to collect network traffic and supplements it with generated data with the characteristics of real traffic.

Keywords: anomaly detection systems, datasets, generative adversarial neural networks, machine learning, computer network security

Stochastic filtering in the space of expert opinions

2022. T.10. № 1. id 1121
Grechanyi S.A.   Krivobokova S.E.  

DOI: 10.26102/2310-6018/2022.36.1.022

Currently, there are many difficulties associated with the financial aspect that affects the security of the facility when assembling a set of technical security equipment. To solve the problems of the facility safety, and therefore the reliability of individual security devices, it is recommended to involve experts. At the same time, experts’ opinions may not always be infallible. In this study, it is proposed to carry out a stochastic analysis and filtering of the expert opinion space in order to identify a conditional trend (established opinion) of each expert. The main method of researching this issue is the analysis of conditional time series (testing the hypothesis of the trend absence by means of the median method in the sequence of expert assessments), which makes it possible to determine the validity of the method, as well as to ensure experts’ objectivity of assessments. The article presents a step-by-step operation algorithm for empirical data, calculates the average deviations from the trend line, considers two methods for testing the null hypothesis – the median method and Foster–Stewart method. The materials of the article can be applied in various areas of setting the average score since the algorithm for testing the null hypothesis is universal in nature.

Keywords: conditional short time series, stochastic filtering, least square method, regression equation, median method, foster–Stewart method, standard deviation

Model with latent parameters for step-by-step procedure for evaluating learning outcomes

2022. T.10. № 1. id 1118
Bratischenko V.V.  

DOI: 10.26102/2310-6018/2022.36.1.015

The relevance of the research is due to the importance of studying learning outcomes to improve the quality of educational process. For this, a knowledge assessment model is proposed in the form of a task sequence. The probability of successful completion of the task depends on the latent parameters: the ability of the student and the difficulty of the task. The model is similar to the Partial Credit Model used in Item Response Theory to analyze test results. In reliance on the maximum likelihood method, a procedure has been developed for estimating parameters by numerical methods according to students' grades. The convergence of the estimation procedure has been substantiated. Adequacy verification of the model by the means of variance analysis, correlation analysis, Infit and Outfit criteria, based on the chi-square distribution, is put forward. To evaluate the usefulness of the model, it is suggested to utilize the coefficient of determination. Information on the application of the model for the analysis of students’ grade array in the academic group is given. Following on from the results of the analysis, the model passed the adequacy tests and made it possible to significantly clarify the characteristics of the learning outcomes and knowledge assessment procedures. To enhance the accuracy of modeling, it is recommended to employ grades of current academic performance. The practical value of the model lies in the identification of assessment procedures with characteristics that differ notably from the average for further meaningful analysis and upgrade.

Keywords: knowledge assessment model, latent parameters, maximum likelihood method, variance analysis, correlation analysis, infit statistics, outfit statistics, coefficient of determination

Modeling of large data array processing in distributed information and telecommunication systems

2022. T.10. № 1. id 1117
Melnikova T.V.   Pitolin M.V.   Preobrazhenskiy Y.P.  

DOI: 10.26102/2310-6018/2022.36.1.025

The article considers the problem of modeling large data array processing in distributed information and telecommunication systems. The analysis of the problem, associated with the incorrect operation of the information system, is conducted. It is shown that distributed systems allow distribution of logic and data across multiple physical servers, which, with the correct allocation of resources and logic, makes it possible to reduce the risks of failures. It is demonstrated how the calculation of productivity loss percentage from a failure in one of the distributed information system components is carried out. A fragment of the diagram, outlining the options for the store information system application, is examined. If it is necessary to purchase physical servers when building an information system, then it is possible to calculate the coefficient of decrease in the performance of the information system in the event of a failure on one of the servers while ensuring the conditions for simultaneous operation of servers. The graph, illustrating the dependence of the performance change on the number of nodes that comprise the distributed network, is given. The analysis indicates that the cases when there are not too many servers have the greatest effect on reducing the risks from adding nodes to a distributed system. If a distributed system already consists of many different servers and any logic is duplicated on many of them, the risk reduction from adding an additional server will be negligible.

Keywords: distributed data processing, information system, modeling, information transfer, algorithm

Detection of defects in faulty elements of power lines using neural networks of YOLO

2021. T.9. № 4. id 1115
Astapova M.A.   Uzdiaev M.Y.  

DOI: 10.26102/2310-6018/2021.35.4.035

Currently, visual state diagnostics of power transmission lines (PTL) elements is a complex and time-consuming procedure. In order to increase the efficiency and reduce the labor costs of this undertaking, the most promising measure is the use of unmanned aerial vehicles equipped with computer vision systems that automatically detect damaged elements of power lines. For the purposes of improving the detection quality of power lines damaged areas by computer vision systems, the application of modern deep neural network architectures would be most effective. However, the problem of utilizing such architectures in the aforementioned task is not sufficiently covered in modern research. The issue of comparing various neural networks and identifying substantial differences in their results is especially acute. This article is devoted to a comparative analysis of modern neural network detectors YOLOv3 and YOLOv4 as well as their reduced versions (YOLOv3-tiny and YOLOv4-tiny) in terms of detecting power transmission line defects. The results of training these detectors on the CPLID dataset are presented along with statistical comparison of the YOLOv3 and YOLOv4 results by means of the cross-validation procedure. The detectors displayed high rates of detection accuracy (mAP @ 0.50 = 0.97 ± 0.03; mAP @ 0.75 = 0.78 ± 0.04) and statistically significant distinctions in these results. A comparative analysis of the findings has revealed that the employment of a simpler neural network YOLOv3 has more potential when applied to detection of power transmission line defects.

Keywords: unmanned aerial vehicle, inspection of high-voltage power lines, fault detection, defect detection, neural networks, YOLOv3, YOLOv4

Optimization of algorithms for the synthesis of controllable systems

2021. T.9. № 4. id 1113
Lomakina L.S.   Mantserov S.A.  

DOI: 10.26102/2310-6018/2021.35.4.040

A generalized probabilistic-structural model and a strategy for identifying the technical state of the system based on the parameters measurement results at specially organized removal points of diagnostic information (control points) are considered. An information measure of the defect localization depth (diagnosis) is proposed which specifies the ratio of information quantity, representing the comprehensive results of the diagnostic experiment, to the information quantity, characterizing the general state of the system. On account of the information criterion, two algorithms for localization of defects in technical systems and technological processes are put forward: an unconditional algorithm, in which testing is performed on a preliminarily selected set of control points, and a conditional algorithm, which implies the choice of each control point depending on the test results from the previous one. The suggested algorithms determine the sequence of control points that ensures the maximum localization depth of the defects and, thereby, facilitating the adaptability of systems to diagnosis, namely, their controllability. In addition, statistical modeling of block failures is assessed, drawing on their priori probabilities, which allows to estimate the amount of information that the test result delivers. The outlined algorithms are stochastic which makes it possible to diagnose complex systems under the conditions of a priori uncertainty, incommensurability of resource (time, performance, memory) and the volume of the problem being solved. Further development of the findings, obtained with regard to accelerated computing and under the conditions of fuzzy information, requires the use of modern information technologies, in particular, neuro-fuzzy modeling.

Keywords: controllability, information criterion, control point, conditional algorithm, unconditional algorithm, statistical modeling

Security analysis of a web application for accessing the critical data storage system

2021. T.9. № 4. id 1112
Vulfin A.M.  

DOI: 10.26102/2310-6018/2021.35.4.038

The paper deals with the issue of providing secure access using a web application to an existing database containing critical information about the parameters of complex technical products life cycle. Based on the analysis of the document of the international organization Web Application Security Consortium (WASC) "The WASC Threat Classification v2.0", possible attacks on a web application, acting as a unidirectional layer of access to the database, exploiting potential vulnerabilities (authentication flaws, authorization flaws, client-side attacks, execution of malicious code on the server-side) have been highlighted and a set of countermeasures has been devised in relation to the architecture of a web application. A pattern has been developed that describes countermeasures concerning the Model-View-Controller architecture of a web application. The diagram of the first level of the web application functional model decomposition is presented. To ensure security at the network level, the basic architecture of the enterprise network with a demilitarized zone and the corresponding configuration of firewalls has been modernized. To assess the security, the internal metrics of software security were utilized, and the cybersecurity risk analysis method by means of fuzzy gray cognitive maps was applied which made it possible to quantitatively assess the reduction with regard to the risk of the accumulated data integrity violation by 3.5 times. Four scenarios of the attacker's impact are considered: without the use of additional countermeasures, the use of the web application layer architectural organization, which takes into account the main patterns of cybersecurity, the use of the Web-application Firewall (WAF), the use of the application architectural organization, and WAF.

Keywords: secure access, basic architecture, model-View-Controller architectural pattern, attack vector, web-application Firewall, fuzzy cognitive map, risk assessment

The application of the relative acoustic nonlinear parameter for the development of biological tissue imaging systems

2022. T.10. № 1. id 1111
Chernov N.   Varenikova A.   Laguta M.  

DOI: 10.26102/2310-6018/2022.36.1.021

The paper discusses the issues of developing a method for visualizing the internal structures of the body, based on the restoration of the acoustic nonlinear parameter distribution. The process of occurrence and propagation of the second harmonic wave in tissues with high nonlinearity and attenuation is considered. The use of the relative acoustic nonlinear parameter in relation to the absolute nonlinear parameter of the medium is proposed. To solve the problem of restoring the distribution of the relative acoustic nonlinear parameter in biological media, an equation is obtained that eliminates the necessity to measure changes in absolute pressure values for both the fundamental frequency wave and its second harmonic. Mathematical expressions are derived that enable accounting for the attenuation processes for the fundamental frequency and for its second harmonic, taking into consideration the influence of the medium in which the object under study is placed. Expressions for determining the relative acoustic nonlinear parameter are acquired. Drawing on these expressions, the construction of a visualization system, utilizing algorithms for restoring the distribution of the acoustic nonlinear parameter in the cross section of a biological object, is presented. The main advantage of these equations is that there is no need to identify changes in the absolute amplitudes of the fundamental frequency and second harmonic waves. The outlined methods for calculating the nonlinear characteristics of biological tissues make it possible to simplify the technical implementation of ultrasound imaging systems.

Keywords: ultrasound imaging, nonlinear parameter, second harmonic, nonlinear acoustics, structure of biological objects

Chatbot based on neural networks and word embedding to increase customer loyalty

2022. T.10. № 2. id 1110
Kovalenko A.V.   Syusyura D.A.   Sharpan M.V.  

DOI: 10.26102/2310-6018/2022.37.2.014

In the digital era mobile devices are becoming the main instrument of human social interaction. With the growing popularity of instant messengers, the role of chatbots in the mobile environment appears to be more and more significant. Intelligent interactive chatbots are often used in mobile applications and help improve the interaction between companies and their customers, which ultimately increases customer loyalty to that organization. Chatbots allow companies to communicate with customers on an individual basis, without involving employees and thereby saving time, money, and human resources. The majority of chatbots works with scripted algorithms and they are not universal. This is due to the simplicity and speed of development. However, in this case, there is a risk of missing many choices in the decision tree. Chatbots based on neural networks can solve this problem, but it should be taken into consideration that both of them have a drawback – long processing of messages and feedback. In the context of the scenario approach, this is caused by long branch transitions. For neural networks, complexity arises because of the feedback processing algorithm. In that instance, the application of the service will not be justified, customer loyalty to the organization will deteriorate. In this connection, the article discusses an alternative approach to creating chatbots with the aid of neural network technologies and text representation methods, which avoids the problems described above. As a means of chatbot design, the following technologies were utilized: Python 3.6, genism libraries, sklearn, scipy, pandas, word2vec and doc2vec technology. The article also describes a way to accelerate chatbot feedback and training using KD-Trees.

Keywords: neural networks, chatbot, word2vec technology, messengers, word embedding

Mathematical modeling of the collective solution accuracy

2022. T.10. № 1. id 1109
Ganicheva A.V.   Ganichev A.V.  

DOI: 10.26102/2310-6018/2022.36.1.001

Currently, the problem of collective decision-making is one of the most relevant in the organization of effective management in social and economic systems. One of the main issues in the theory of expert assessments is the assessment of the group solution quality. The article discusses the matters of assessing the socio-economic indicator by independent experts. The centered random variables sums value of individual estimates is accepted as the error of group estimation. The situation is examined when the values of the indicator have an arbitrary distribution with known and unknown parameters. Two algorithms have been developed to determine the required amount of experts depending on the accuracy and reliability of the assessment. The first algorithm is used to find the confidence interval of mathematical expectation when the variance of the indicator is not specified. In this event, an iterative process is undertaken to ascertain the volume of representativeness for the confidence interval of variance with a given accuracy and reliability. The second algorithm is employed to construct a confidence interval for variance when the number of experts is more than three. The important task of quantifying the proportion (percentage) of possible errors within a predefined interval in measuring the indicator has been solved. An econometric model is designed for the Laplace function. The case of determining the number of experts to evaluate an indicator having a uniform and exponential distribution over a given interval is considered. An example of the practical implementation of the devised method is shown.

Keywords: approximation, significance level, estimation accuracy, laplace function, econometric model, expert, estimation, probability distribution

Implementation of the method for dynamic content matching of learning and game scenarios in an adaptive learning game

2022. T.10. № 1. id 1107
Khairov A.V.   Shabalina O.A.   Kataev A.V.  

DOI: 10.26102/2310-6018/2022.36.1.002

The paper describes a method for adapting the educational process in adaptive role-playing learning games using dynamic content matching of nonlinear learning and game scenarios. The training scenario and the game scenario, associated with it, are represented by structurally ordered spaces. The mechanism of dynamic content matching of nonlinear training and game scenarios implemented in the game is based on the activation of predefined game events in relation to the plot path chosen by the player. Triggers are assigned to each visible and invisible object of the game world in conformity with the constructed knowledge space. When a character enters a game situation tied to the trigger's scope, a system of dialog boxes is enabled to complete the objective, set by this game situation. The player dynamically builds their game scenario depending on the choice of a strategy for mastering the knowledge space and performing the corresponding game tasks while the method of dynamic content matching, executed in the game, ensures the development of the entire space in terms of any strategy formed by the player. The means of the method employment in the Cammi learning role-playing game for studying object-oriented programming and the C++ language is defined.

Keywords: learning game, adaptive learning game, training course, adaptation model, adaptation method, nonlinear scenario, knowledge space

Risk analysis of social media content based on neural network classification of a message text emotional coloring

2021. T.9. № 4. id 1105
Razinkin K.A.   Sokolova E.S.   Savishchenko D.N.   Chapurin E.Y.  

DOI: 10.26102/2310-6018/2021.35.4.034

One of the promising areas of Data Science within the framework of practice-oriented approaches to the analysis of social networks (Social network analysis) from the point of view of network users’ (agents’) opinion formalization is a class of content analysis methods designed for automated identification of emotionally colored vocabulary in texts and emotional evaluation of authors in relation to the objects referred to in the text. With the help of such an analysis, it is possible to study an array of messages and other data and determine how they are emotionally colored - positively, negatively or neutrally. The article offers a comparative analysis of two approaches to the study of text sequences classification possibilities depending on their emotional coloring: one by means of a recurrent neural network (RNN) and another involving graph convolutional networks (GCN). The first approach is implemented through deep learning utilizing the Deep Learning Designer tool (MathWorks © MATLAB R2021b). The second approach is based on the application of convolutional graph neural networks for text classification. GCN implementation is carried out in Python using the appropriate set of libraries for data analysis. In addition, the paper shows that the resulting model can be used in risk assessment, where the resulting value serves as a correction factor in calculating the risk of user involvement. Based on the results of the two approaches comparison, it is shown that when using GCN, the percentage of training data decreases, which indicates the sensitivity of the method to a smaller amount of training data, while the accuracy of the model increases with comparable configurable training parameters

Keywords: emotional coloring of the text, recurrent neural network, deep learning, graph convolutional networks, risk analysis

Choice optimization of the purposeful team activity tasks structure in an Agile-oriented organizational system

2021. T.9. № 4. id 1104
Borzova A.S.   Korchagin S.G.   Lvovich Y.E.  

DOI: 10.26102/2310-6018/2021.35.4.039

The paper proposes to select the structure of the purposeful team activities tasks in an Agile-oriented organizational system based on the optimization approach. The need for optimization is justified by the redundancy, ambiguity in traditional expert assessment methods of the goal achievement process at a given time for ensuring the values of the indicators set by the control center. The implementation of the optimization approach is fulfilled within the framework of modeling and double reduction algorithmizing of the initial set of problems, defined at the expert level. At the first stage, the application of multi-alternative optimization and quantitative expert assessment is carried out to form a numbering set of tasks, facilitating the achievement of the team activity goal. At the second stage, it is employed to decide on the sequence of tasks that are included in the reduction set when organizing the iterative process adopted in Agile-oriented organizational systems. To make management decisions, a step-by-step randomized search scheme is used that combines both stages of double reduction. The final selection is made in reliance on a set of expert rules. As a result, we obtain a variant of the task structure that enables, within a given time frame and with a certain number of team members, the goal attainment with the specified values of quantitative indicators.

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

Mathematical modeling of hydrogen extraction process from the natural gas steam reforming products

2022. T.10. № 1. id 1103
Alruyshid M.H.   Skvortsov S.A.   Ishin A.A.   Dmitrievsky B.S.   Terekhova A.A.  

DOI: 10.26102/2310-6018/2022.36.1.024

Nowadays, the problems of the economy decarbonization and the transition to renewable energy sources are becoming more relevant. Hydrogen, above all, is among the decarbonated gases, the use of which is considered the most promising. It is regarded not only as an energy carrier but also as a means of storing excess energy, produced by renewable sources. The steam methane reforming, at the outlet of which a hydrogen-enriched gas stream is obtained, is deemed to be the most well-developed scheme. Adsorptive gas separation systems are widely used to extract hydrogen from the steam reforming gas. Mathematical modeling plays an important role in the design of adsorption gas separation plants. Optimization and control systems synthesis can be carried out with the help of a mathematical model for the process under review. The article proposes a mathematical model for the hydrogen extraction technological process by the pressure swing adsorption method employing a 6-beds gas separation unit. For the numerical solution of the mathematical model equations, the method of lines was applied, which converts partial differential equations to a system of ordinary differential equations. The integration of resulting equations system was undertaken with the aid of the Runge-Kutta method with automatic step selection. The results of numerical simulation calculations of the system dynamic operating modes are given. It is concluded that the proposed mathematical model for the adsorption process of gas separation is allowed for solving the problems of numerical simulation studies, optimization and control systems synthesis

Keywords: adsorbent, adsorber, adsorption, activated carbon, regeneration, zeolite

Analysis of methods for solving inverse kinematics of modular reconfigurable systems

2021. T.9. № 4. id 1101
Erashov A.A.   Blinov D.V.   Saveliev A.I.  

DOI: 10.26102/2310-6018/2021.35.4.025

The relevance of this work is due to the actualization of methods for solving the inverse kinematics in relation to various kinematic structures (formations) of reconfigurable modular systems. The purpose of the work is to analyze methods for solving the inverse kinematics, which can be applied to various formations of self-configuring multilink robotic systems. A study of the forward kinematics of modular robotic systems various formations is conducted on the basis of the previously obtained research results of other scientists. The analysis of methods for solving the inverse kinematics of modular reconfigurable systems was carried out and an assessment of their possible application for various kinematic structures of modular systems was made. Analytical and numerical methods of solution were considered, and examples of practical application were also given. In addition, the paper analyzed various machine learning methods. With regard to the results of the study, the advantages and disadvantages of various methods for solving the inverse kinematics of modular robotic systems were highlighted. Potentially suitable methods for solving this problem from the point of view of computational complexity and application possibilities for systems with a redundant number of degrees of freedom are identified. Among the methods considered, particular solutions of the inverse kinematics of a certain modular reconfigurable system kinematic structure are often evaluated. As a result of the analysis, it is possible to isolate areas of research related to the development of machine learning methods that are potentially suitable for use in control problems for self-reconfiguring modular robotic systems. The development of such a method will enable to reduce the number of preliminary analytical calculations, to implement a control system that does not require significant changes in algorithms, and also to expand the possibilities of using modular systems by adapting this system to the movement surface.

Keywords: modular robotics, modular robotic systems, self-reconfigurable modular robots, autonomous robots, forward kinematics, inverse kinematics