This PDF file contains the front matter associated with SPIE Proceedings Volume 13065, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

This article describes the creation of a custom thematic dataset for generating descriptions of artistic works (paintings). An algorithm for forming a dataset using deep learning models is proposed. Initially, an analysis of scientific publications was conducted to identify key features important for the perception of works of art. Then, relevant datasets with descriptions in various languages were extracted from various sources, including the collections of major museums. After preliminary processing and filtering of irrelevant features, the data were combined into a single dataset. All text data were translated into Russian. 7 key features and 21 fields were selected for each painting image with corresponding descriptive information. As a result, a new dataset consisting of 10,000 images of paintings and annotations to them was formed.

Protocols play an important role in decision-making across many fields. The effectiveness of the work of various organizations and teams directly depends on their quality and speed of writing. Therefore, automating this process is of utmost importance today. This article proposes an approach to address the issue of logging by processing existing audio recordings of meetings or events and using an ensemble of artificial intelligence models, including a pre-trained acoustic speech recognition model based on the "Quartznet" architecture, a language N-gram model implemented using the "KenLM" tool, and a Russian-language model "RuBERT" retrained for extractive abstracting. The algorithms of these models' operation are examined, and the processes of data collecting and retraining, as well as the implementation of the selected models, are described. Quality metrics for the speech recognition system are compared, and an analysis of the implemented systems is conducted. To utilize the developed system an HTTP web server was deployed via an API. The final results of the developed automatic logging system, capable of extracting timestamps of spoken words, highlighting speakers in audio recordings, and reducing the resulting text to a specified percentage are demonstrated.

In the field of automation and telemechanics, the main task is to improve the automatic locomotive signaling system (ALS), including the creation of a complex of remote measuring devices using wireless, microprocessor and digital technologies to eliminate malfunctions on railway sections of automatic and semi-automatic locking. This article presents an analysis of errors in traffic light indicators and the automatic locomotive signaling system by month from the time of the year, which affect the speed of trains, namely, due to a false indicator, the speed decreases, this leads to lagging behind schedule, thereby due to the lack of timely arrival of trains, the company incurs financial losses. All malfunctions arising from a system failure are represented by mathematical expressions, on the basis of which recommendations for elimination are given. A module for receiving information from the locomotive's ALS was developed based on modern wireless technologies, and for the accuracy of checking the system, a stand based on the locomotive's ALS blocks for measuring signals was developed, which corresponded to the exact readings of the parameters of the ALS code signals. The developed module provides an opportunity to instantly notify the site's electromechanics of identified deficiencies in the ALS system.

The purpose of this article is to develop an information-measuring complex for studying the influence of mechanical stresses in metal on the parameters of electromagnetic-acoustic conversion. Based on a comparative analysis of existing installations for studying the acoustic and electrophysical properties of metals, the structure and composition of the information-measuring complex have been developed. When developing the complex and planning research, the requirements of current regulatory and technical documents were taken into account. The research results are aimed at improving methods for isolating and processing information contained in the signal of an electromagnetic-acoustic transducer to increase the sensitivity and information content of non-destructive testing and diagnostics of metal structures.

The study is devoted to the possibility of using neural network algorithms to monitor the current and voltage signals parameters during transients in electrical engineering complexes. The direct propagation neural network can be programmed in standard microprocessor equipment and allows approximating signals in real time in the "sliding time window" mode. The proposed approach is demonstrated in laboratory experiments on the asynchronous motors run-out physical model as a complex load node part. It is shown that the multilayer perceptron use makes it possible to track changes in the signals frequency and amplitude with percentage accuracy for the observation time window duration of no more than a industrial frequency quarter period. The results demonstrate the using neural networks advantages for signal processing in non-stationary modes in electrical engineering and electric power industry.

The influence of the degree of blur on the recognition of high-contrast simple objects is studied using images of handwritten digits and letters as an example. High-contrast objects are objects that are significantly brighter than the surrounding background. White images of handwritten digits (MNIST) and Latin letters (EMNIST) against a black background measuring 28 by 28 pixels are selected as training and testing data sets. It is found that recognition accuracy decreases linearly as the blur level increases. The work also shows that both an increase of blur and a decrease of blur from the blur level of the training sample worsens recognition. It is concluded that recognition when blur is reduced is worse than when blur is increased. Histograms are presented that demonstrate the dependence of recognition accuracy on the degree of blur. It is shown that the initial weights affect the recognition accuracy.

The paper presents the results of research on the implementation of password protection as a way to ensure information security. The varieties of passwords used in practice (user-set passwords, system-generated passwords, half-words, key phrases, interactive “question-answer” sequences) are considered, their advantages and disadvantages are analyzed with illustrative examples. Taking into account the current regulatory documents, recommendations on the formation of strong user passwords and actions of information security specialists to maintain the relevance and functionality of password protection are proposed. Using the high-level programming language C++, various variants of the password verification process that can be implemented in corporate information systems are discussed in detail. In addition to full source codes of computer programs, detailed comments on the logic of their functioning are given.

System analysis of multidimensional random data is one of the most important tasks of modern information technologies. Traditional statistical methods for analyzing such data do not always allow one to obtain adequate results. In this article, the joint use of the decision tree method and the artificial neural networks is proposed for data classification. It is shown that this approach can be an alternative to multiple factorization of multidimensional data and has a number of significant advantages over traditional statistical methods. Even when the data has limited and different digital scales, the use of machine learning makes it possible to conduct a hierarchical classification of intra-system data relationships, to reduce the number of parameters significant for analysis, and significantly reduce the size of the learning sample required. For the technical applications, this allows for real-time data analysis using a microprocessor implementation of neural network algorithms.

The problems of rational use of natural resources and subsoil, in particular, are in the sphere of constant attention both from state bodies and from subsoil using enterprises. The data about the subsoil is enlarged by conducting geological exploration, subsequent interpretation of their results, submission of reports and other accompanying documents to the relevant geological information funds. However, the accumulation of primary and interpreted geological information does not contribute to the formation of a unified knowledge base that could allow to prepare analytical information automatically to support management decision-making. To eliminate this problem, it is advisable to use an ontological approach. In this case it is connected with the construction of cause-effect relations between elements (entities) and filling the resulting ontological model with factual information about the geological indicators obtained as a result of the work carried out. The task of using automated information systems to assess the efficiency of subsoil areas is also becoming very relevant. This is due to the increasing role of expert information and analytical systems in the structure of software products used for analyzing the results of geological exploration and determining new priorities in the research of mineral resources of solid minerals and groundwater for the needs of the region. The decrease in the number of geologists in this field leads to the need to design a system of geological assessment of the prospects of subsoil areas based on an ontological approach. The application of ontology is aimed to reproduce the chain of decision-making by an expert, but in a new view based on associative relations and empirical rules.

The integrating role of the transport schedule of regular passenger transportation for the tasks of the tactical and operational levels of transport system management determined the relevance of the presented study. The purpose of the study is related to the development of methods and algorithms for the formation of schedules of regular passenger transport. The formulation and formalization of the task using the schemes of formation of schedules and priority rules are carried out. For numerical experimentation, a mock-up of a program for forming a railway schedule and a test task were developed, including 4573 applications for weekly schedules of a railway network containing 100 stations, 128 runs between stations, 100 passenger routes of different frequency, 471 trains per week. The developed layout of the program, in addition to performing calculations using the presented algorithms, visualizes the results of numerical experimentation in the form of spiral representations of station schedules and runs. The operability of the developed methods and algorithms has been confirmed. The conclusions and discussion of the results of the study are presented.

The article explores ways to geocode events and extract information about the location where the specified event occurred. A tool has been developed that allows you to extract information such as addresses and intersections from text messages, create datasets and perform geolocation for both historical datasets and streaming data. Social networks were used as data sources, where the user can select a range of communities to search for information. The entire process of searching for messages, processing texts and placing geotags on the map is fully automated.

The goal of this research is to diagnose breast cancer using machine learning methods including decision trees, support vector machines (SVM), and naïve Bayesian classifiers. The properties of cell nuclei taken from breast biopsies are included in the Breast Cancer Wisconsin dataset, which is used in this study. Three different machine learning algorithms - naive Bayesian classifier, SVM and decision tree - are used to develop predictive models aimed at classifying samples as malignant or benign tumours. The study involves training the models on evaluation data and then evaluating their performance on test data. In order to compare the effectiveness of each method, many metrics like accuracy, sensitivity, and specificity are used in the evaluation of the results. The outcomes demonstrate the efficacy of every technique examined in this research. This work can be used as a springboard for future research into refining and customizing these techniques to intricate clinical situations, in addition to offering a useful comparison of the efficacy of three distinct machine learning approaches for breast cancer diagnosis. The results may facilitate the application of machine learning techniques in clinical settings, hence facilitating early detection and bettering the prognosis of patients with breast cancer.

The article presents the results of using microscopic simulation modeling to improve the operation of road infrastructure and optimize the phases of traffic light regulation on an emergency-prone section of a highway in a city with a population of more than 500 thousand people. Based on the results of field studies, a simulation model of this site was built. Using statistical criteria, the high validity of the developed computer model was confirmed, which ensures the reliability of the results obtained. In the process of multiple virtual experiments, the optimal mode of operation of the traffic light was determined, which ensures a reduction in travel time, an increase in the throughput of the analyzed section of the road, as well as a reduction in emissions of pollutants from vehicle exhaust gases.

This article analyzes the communication protocols and architecture of multi-agent systems. In the introduction, the general purpose of multi-agent systems and the principles of information exchange are described. Annotated with examples and graphs to analyze communication protocols and architectures. The article provides information on the structure, characteristics and areas of use of communication protocols. Also architectural models (SOAA, BDI, Gaia) were comparatively analyzed. This article provides information related to a specific topic of interest to professionals and industry researchers, system administrators.

The article considers the main stages of designing digital clothing collections, software products used at various stages of creating digital clothing, suggests ways to digitally design various patterns on fabrics, as well as the technology of digital clothing design when designing the appearance of a product, digital clothing models have been developed taking into account various color and texture design of materials.

Various stochastic optimization methods are utilized for training neural networks. The objective of this research is to provide a comprehensive overview of stochastic optimization methods proposed to enhance and expedite the convergence of neural networks. The article presents a highlighting of stochastic optimization methods' advantages and drawbacks, while also analyzing the constraints of their applicability. It commences with an introduction to the problem formulation, followed by sections dedicated to various algorithmic modifications: SGD-based stochastic optimization methods, adaptive gradient methods, and methods of adaptive moment estimation. In conclusion, the article underscores the importance of a judicious selection of a method, contingent upon its characteristics, applicability constraints, specific task, model architecture, and data quality.

The "covtype" dataset in scikit-learn represents forest cover information and includes a variety of soil characteristics for seven different forest cover types. The proposed work solves the classification problem, where the goal is to accurately determine the type of forest cover based on given soil characteristics. The study uses various machine learning methods such as decision trees and naive Bayes classifier. Models are trained on an extensive training set and then evaluated on test data to determine their ability to accurately predict forest cover types. The classification results are analyzed, including metrics of accuracy, recall, F1-measures, as well as ROC curves are constructed and the areas under them (AUC) are calculated. The results and metrics obtained allow us to compare the effectiveness of different models in solving a given classification problem. The knowledge gained can be useful for the application of machine learning algorithms in ecology and forest resource management.

In the presented work, the method of localization of vulnerable sections of the centralized water supply network using a neural network approach is considered. This method will allow you to identify areas where there is significant wear on pipes, leading to emergencies and water leaks. The paper discusses existing methods for assessing water quality using artificial intelligence. The adaptive neuro-fuzzy network ANFIS was used in the work. The input data are electrical conductivity, ppm, hardness, iron and flow rate in pipes of various diameters. The output value is the pipe wear value. The ANFIS network model alternately applied two types of training: back propagation and a hybrid method, aggregating backpropagation and least squares. The Fuzzy Logic Toolbox alternately applied membership functions: piecewise linear, Gaussian distribution, sigmoid curve, quadratic and cubic curves. A comparative analysis of the errors of different types of membership functions and learning algorithms was carried out. The coefficient of determination for the simulated and initial data is determined. The research was carried out on a section of the centralized water supply network, one of the districts of the city of Murom, Vladimir region, Russia.

In recent years, the integration of machine learning algorithms has significantly advanced the field of computer vision, particularly in the domain of object detection and tracking in video images. This article explores the application of machine learning techniques to enhance the accuracy and efficiency of object detection and tracking systems. The article begins by providing an overview of the challenges associated with traditional methods of object detection and tracking, highlighting the limitations in handling complex scenarios and diverse object types. Subsequently, it delves into the methodology of employing machine learning algorithms, emphasizing their adaptability and capability to discern patterns and features crucial for accurate object recognition. Various machine learning models, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), are discussed in detail, elucidating their roles in extracting meaningful representations from video frames. The training process is explored, encompassing the use of labeled datasets to enable the algorithms to generalize and make informed predictions in real-world scenarios. Furthermore, the article investigates the integration of deep learning techniques, exploring the advantages of transfer learning and fine-tuning pre-trained models to optimize performance. The role of neural networks in handling object occlusion, scale variations, and pose changes is emphasized, showcasing the adaptability of machine learning algorithms to dynamic and unpredictable environments. The practical implementation of these algorithms in object detection and tracking systems is presented, highlighting real-world applications across industries such as surveillance, autonomous vehicles, and industrial automation. The article concludes by discussing ongoing research and potential future developments, addressing the evolving landscape of machine learning in computer vision and its implications for advancing object detection and tracking capabilities in video images.

In the article describe the methods of increasing the contrast of images in video information systems, the main compression of the video stream is provided by eliminating inter-frame redundancy using motion compensation methods for image fragments of adjacent frames. However, the use of motion compensation methods requires the formation of additional data (metadata) containing information about the types of image blocks used, the coordinates of their movement, etc. At the same time, in order to increase the compression of the video stream without compromising its quality, a higher accuracy of motion compensation is required, which leads to an increase in the number of blocks and, accordingly, to an increase in the volume of metadata that reduces the effectiveness of motion compensation. This is the main problem of compressing streaming video without degrading the quality of images. In addition, the higher accuracy of positioning blocks with motion compensation dramatically reduces the speed of image processing, which is not always feasible in real-time systems.

The work defines the features of financing the company's commercial projects, taking into account the formulated additional functions of financing based on the possible variability of alternatives for the long-term management of the company. The possibility of using a decomposition approach when choosing a solution to a global problem is substantiated, as well as an algorithm for generating a complete set of target indicators that satisfy specified restrictions at the lower hierarchical level. The possibility of using elements of the graph approach in the procedure for comparing budget options is substantiated, and statements are formulated that guarantee the reliability of preference relations for a specific criterion.

This article presents a comparison of the effectiveness of using different architectures of neural networks in solving the problem of detecting anomalies in time series. The performance characteristics of data storage systems as a subject area are presented. Data is taken from storage performance logs. The article compares the effectiveness of LSTM, SVM, Autoencoder, CNN, RNN architectures, and also presents the developed LSTM architecture. This neural network shows the highest results as a result of experiments. In the future, it is planned to refine the algorithm by interpreting the results using semantic knowledge bases based on ontologies and rules.

This article presents a description of a software service for analyzing text information for the formation of an OWL ontology. This software service analyzes the state of complex technical systems. The work describes the principles of operation of the service and presents the results of experiments on preprocessing and clustering of text data. The experimental results confirm the possibility of using the service to generate OWL ontologies.

Currently, one of the most vital issues concerns dealing with smart contracts is security. Concerns over inefficiency, security, and fraud are grounded, and ignoring them when using the blockchain network to create smart contracts can result in extremely high additional costs. This article covers the most common tools for auditing smart contracts. The purpose of this article is to compare the most popular software tools to analyze smart contract security. For this purpose, the following types of smart contract vulnerabilities were investigated. This article provides examples of the major smart contract vulnerabilities, and given the relevance of digital agreements and the catastrophic consequences of errors, recommendations have been offered to redress smart contract vulnerabilities.

The development of any modern communication system is impossible without an accurate understanding of the channel that is supposed to be used to transmit information. At the same time, at the stage of theoretical research it is necessary to use mathematical models of channels, the adequacy of which to the real processes occurring in radio links largely determines the validity of the technical decisions made. This article is devoted to modeling modern MIMO systems in channels with various fading conditions. For this purpose, mathematical models of continuous communication channels, represented in the form of stochastic differential and difference equations, in radio relay and tropospheric communication lines were obtained and analyzed.

This article explores the important issue of protecting software from reverse engineering techniques by presenting an approach based on the use of an encrypted command stream interpreter. This article proposes the use of an encrypted command stream interpreter as a security measure. This method complicates the process of code analysis and makes it less vulnerable to attacker intervention. The article describes an approach to implementing and using an encrypted command stream interpreter, emphasizing its applicability in various fields, including information security and software protection. The presented method provides protection against reverse engineering by encrypting and obfuscation of code, which is a contribution to the field of cybersecurity.

The article considers autonomous vehicle control system using obstacle detection method based on machine vision using light markers on motion path. The data gaining method for optimization and calculating image is proposed. The specific feature of realizing the aim is light markers grid deformation consideration. The software realization of obstacle detection based on light markers grid geometric deformation is demonstrated.

This article establishes the knowledge base for the decision-making module within the information monitoring system software package designed for the continuous production stages of oil and fat enterprise. The construction of the knowledge base relies on a frame model, which serves as a mathematical framework for the conceptual representation of objects. Object-oriented programming in C# is employed to implement the knowledge mechanism outlined by the frame model. The knowledge base is formed through abstract frames and the objects derived from them, specifically tailored for the storage and processing of oilseeds. Decision-making processes are then executed based on the information stored within this knowledge base.

The issues of correctness of using the BRISQUE metric for assessing image sharpness are addressed. For this purpose, groups of images of three sizes, namely 150x150, 300x300 and 500x500, are selected. These images are deteriorated by horizontal blur and then convoluted with a 3x3 convolution matrix. The matrices are chosen such that their elements surrounding the central element have identical values. Blurred images are reconstructed by such convolution matrices, with elements varying over a wide range. All reconstructed images are evaluated using the BRISQUE metric. In the case of specific images, graphs of the dependences of the BRISQUE values on the values of matrix elements are presented. The extrema of the BRISQUE graphs are shown. It is shown that the BRISQUE can take negative values and its minima may not correspond to sharp images.

A practical case for implementing software systems for emulating 2G/4G mobile communication systems, based on the use of Osmocom and srsRAN 4G software packages, is considered. The main functions and capabilities of software packages are described, as well as the features of their use in studying the functional characteristics of radio access network elements, transport and core networks is considered.

The design and management of a product knowledge base (PKB) brings forward important aspects in the field of information technology. This article discusses strategies for creating and effectively managing a system based on production rules. The theoretical and practical foundations of PBP design are analyzed, with attention paid to the generation of knowledge from voluminous information data. The paper proposes an algorithm for structuring a subject area through its description using knowledge, which serves as the basis for creating a production knowledge base. The paper highlights the issue of the need to develop algorithms for determining information sets with cause and effect properties, aimed at eliminating subjectivity in the formation of the subject area and information sets. The process of formalizing the concepts of “data” and “information” builds an analytical definition of “knowledge”. In this work, based on production knowledge, the concept of “fractality of knowledge” is introduced, and an algorithm for designing a fractal production knowledge base (FPKB) is proposed. The control algorithm of the FPBZ is also presented, and the decision-making procedure is based on the evaluation algorithm (ABO), designed to give the system universality in the sense of its functionality.

This work is devoted to the study and application of ensemble methods in the problem of classifying forest cover types based on the "covtype" data set. The paper examines two popular ensemble methods: random forest and gradient boosting . First, data analysis and preprocessing is carried out, including dividing the sample into training and test sets. Then random forest and gradient boosting models are built on the training set. F1-measures, as well as ROC AUC. Results of study shows that both ensemble methods effectively cope with the task of classifying forest cover types. The resulting metrics confirm the high accuracy and ability of the models to generalize to new data. An important step in the research is to compare the performance of random forest and gradient boosting . The work also includes visualization of results such as ROC curves for further exploration and comparison of the two methods. The findings can be useful for choosing the best method in specific scenarios and understanding their applicability in natural data classification problems.

The method of creating electronic cards of agriculture has been improved, taking into account the classification and content elements of GAT and agricultural cards. With the help of GAT technology programs, a technological system for creating and integrating an agro-geodatabase was developed, and the method of visualizing agricultural data was improved. Agricultural cards were classified according to their creation, purpose, territorial coverage, equipment, scale, method of use and content, and the content elements of the electronic card were developed. Based on GAT, a technological system for creating and converting an agro-geodatabase was developed, and interactive and web maps of agriculture were created.

Training to initial device configuration and use of Packet Tracer network simulation software are reviewed in the article. The various types of router and switch interfaces are described and the main methods of accessing network devices through the console port are shown. The authors emphasize that knowledge of initial device configuration and packet switching network modeling is essential for information technology professionals. Use of Packet Tracer allows to master own skills in networks creation, check the correctness of settings and determine the causes of possible problems in the network. The authors recommend to use the Packet Tracer software for training and practicing skills in setting up and configuring the devices in packet switching networks.

The purpose of this article is to develop methods for representing scientific documentation in the form of graphs and their subsequent use. It analyzes the current state of quantitative and bibliometric methods for document analysis and possible ways of their application. The article also describes the developed software that serves to visualize the analysis of the conceptual structure of scientific publications and what progress has been made during the study. The result of the work carried out is a software product oriented to use during training at universities, with the ability to view scientific publications in the form of graphs. The software product itself is presented in the form of a web page, with built-in functionality for working with scientific materials and a database where the processed information is stored.

This article is devoted to the consideration of how information technologies are used in practice in the agro-industrial complex at the present time. The main purpose of the study is to substantiate the effectiveness of these methods and to identify the main areas of use. The paper discusses the use of machine learning models to work with data in agriculture. The relevance of the study is justified by the fact that industry, as well as agriculture, are one of the most priority areas in our country. There is a need for the use of adaptive devices and new methods that will make it possible to achieve significant success in this area. The article is devoted to the consideration of the code implemented in Python, which can predict crop yields with high accuracy. By applying adaptive learning models and comparing different statistical models with each other, it is possible to identify the most accurate predictive indicators.

To date, the problem of automating work with images taken using satellite systems has become relevant. This task concerns a wide range of human activities, including urban planning, transport logistics, ecology and environmental monitoring, etc. To solve these problems, there are many tools, of which solutions based on the use of machine learning algorithms are particularly effective. The complexity of this approach lies in the wide variety of computer vision models that exist today. The purpose of this research работы is to select the most popular neural network architectures and conduct a study that aims to identify the most effective architecture in terms of efficiency and quality of work performed. This study will help determine the machine learning model that is most suitable for further use in a software product aimed at working with satellite images, the main functions of which will be object detection and segmentation.

The article is devoted to the problems of using digital platforms in systems for managing industrial companies. The purpose of the article is to study digital platforms in the management systems of industrial companies. The research materials of the article include basic scientific and practical information on the use of digital platforms in the management systems of industrial companies. The research methods of the article include the systematization and generalization of data on the topics of modern trends in the use of digital platforms in the management systems of industrial companies. The paper highlights the main aspects of the introduction of digital platforms in the activities of modern companies. It is shown that the digital platform differs from other software products in that its design is focused on a fundamentally wider range of services. In those areas of consumption where it is possible to create a single information space, and the service or product itself is unified, it is very difficult to compete with digital platforms and structures built on a network principle. The functioning of digital platforms for managing industrial companies is not limited by borders, is not based on a territorial principle, they can work anywhere. The physical location of servers and software required for the operation of digital platforms for the management of industrial companies does not matter for the digital society, since thanks to the Internet, it can work anywhere. Everything that is not based on complex technology is more profitable to organize on the basis of platforms. A model of a product digital platform for an industrial company was developed, which reproduces the life cycle of its industrial products developed for market customers and released for free sale.

The aim of this study is to evaluate the methods of autonomous collection and use of local contexts for selecting a response option by an intelligent system in the process of dialog interaction. The author analyzes the methods of designing interactive intelligent systems and identifies the main technologies used to implement mechanisms for recognizing user requests and selecting response options. The following research methods were used: technical documentation, scientific papers, conference materials and methodological manuals. The current methods of collecting local contexts used in the most successful projects are analyzed. Methods of encoding text words for software processing are analyzed. Methods of measuring the quality of data collected in this way are analyzed. The result of the work is a presentation of the studied methods for collecting, preserving, and evaluating local contexts. A scheme for data exchange between modules that implement this functionality is also proposed. The results of the work can be used to create a system that combines these methods to improve the quality of dialog interaction and implement self-regulation based on evaluating the power of collected thesaurus. It is shown how it is possible to obtain all the data necessary for the correct functioning of each of the methods within a single algorithm. A comparative analysis of the complexity and effectiveness of individual methods for collecting local contexts and the proposed integrated system is made. The data collected in this way preserves logical connections with the texts that surround them and can be effectively used both for linguistic research aimed at studying the unique features of the use of words in various fields of activity, and for improving the quality of the choice of responses of dialog interaction systems.

This article proposes a new approach to predicting failures of high-tech manufacturing equipment using machine learning methods. As part of the study, various modern approaches for solving similar problems were analyzed, their advantages and disadvantages relative to the proposed method were described. Model training data includes sensor readings measuring temperatures, pressures, currents, and vibrations for six exhausters over three years. In the process of data preprocessing, standardization, noise removal and extraction of new features were carried out. Since blower failure is a rare and short-term occurrence, the use of supervised learning methods is difficult. In this regard, it is necessary to introduce an additional stage of screening out periods of normal operation. Thus, the process of early diagnosis of failures consisted of two successive stages. Firstly, anomalous operating modes of equipment were identified using the isolation forest method. Next, intervals were classified using a recurrent neural network, defined as anomalous, according to the technical locations that caused the equipment failure. As a result, on the test sample it was possible to achieve an average value of the F1 metric equal to 0.74 for all analyzed technical locations.

The article deals with the problem of mathematical modeling of physical processes when conducting diagnostic tests within the framework of non-destructive testing of materials. The consideration of elastic processes is due to the fact that elasticity is characterized by the return of the body to its original state after removal of force loads (thereby ensuring the non-destructive nature of the tests). It is assumed that the material under study has weakly expressed spatially distributed anisotropic and nonlinear properties due to the uneven accumulation of microdamage during the operation of the product. The purpose of the work is to formulate and solve the problem of diagnosing a nonlinear elastic inhomogeneous anisotropic half-space within the framework of mathematical modeling of non-destructive testing of massive bodies. In mathematical terms, the problem belongs to the class of coefficient inverse problems for dynamic elasticity equations. An approach to determining density and elastic moduli (coefficients of equations) depending on spatial variables is proposed, based on the use of the perturbation method.

This article presents an algorithm for determining reference brightness correction coefficients to improve image quality. The algorithm utilizes a combination of statistical analysis and image processing techniques to identify and correct brightness discrepancies in digital images. By establishing a robust and efficient method for calculating these coefficients, the algorithm aims to enhance the overall visual fidelity of images across various applications, such as photography, medical imaging, and remote sensing. The proposed algorithm demonstrates its effectiveness through experimental results and comparisons with existing methods, highlighting its potential for practical implementation in image processing workflows. In the world, the filtering of digital images by the convolution method with a pulse characteristic in the spectral region scientific research is being conducted to improve the quality level of digital television images, methods for modeling filtration processes and highly efficient control systems in a number of priority areas, including: on the formation of mathematical models of filtration processes, improving the methods of wavelet, Fourier, Haar, Walsh-Hadamard, Karhunen-Loev in increasing the clarity and brightness of images based on linear and nonlinear differential equations; creation of methods for eliminating additive, pulsed and adaptive-Gaussian types of noise in images using additive and adaptive filtering; methods of algorithms and software for introducing intra-frame and interframe image transformations; methods of adaptive brightness system control using the Chebyshev matrix series; methods of gradient, static and Laplace methods for image segmentation and dividing it into contours; formation of criteria and conditions for evaluating image quality. Conducting scientific research in the above research areas confirms the relevance of the topic of this article.

This article explores the key components for enhancing ETL performance, focusing on multi-threaded data loading. The research begins with a survey of existing ETL solutions to identify their strengths and limitations. The results of this research underscore the significance of incorporating multi-threading and flexible loading parameters into ETL workflows. The article demonstrates that by leveraging multi-threading and selecting an optimal data loading package size ETL processes can achieve a remarkable speed improvement, almost tripling the data loading speed.

The article discusses intelligent ways to regulate room temperature. It is shown how the project architecture is built and the tools for its implementation are selected using intelligent algorithms, mainly fuzzy logic. Recommendations are given for the selection of software and hardware for implementing microclimate control systems. The hardware can be implemented using a NodeMCU board with a built-in ESP8266 module, and DS18B20 integrated temperature sensors connected to it. It is recommended to program the device in the Arduino IDE. The software part can also be implemented as a web service and a database connected to it. The web service is implemented using the Flask framework. The database is implemented in the MySQL environment. Nginx is selected as the web server for deploying the web service. The results of this work can be used in various rooms with installed heating, air conditioning and ventilation devices to maintain optimal microclimate conditions with minimal energy costs.

This article provides formulated information about the design and implementation of the information system, the software product "Information system for personnel management of the department "Ecology and labor protection"- (e-pedagog.uz)" was obtained, which allows automating the activities of the department of a higher educational institution. The software was fully tested in accordance with the program and test methodology. During testing, a number of errors regarding the correctness of calculations were identified and corrected. The check showed the operability of the system on all declared operating systems, as well as on various hardware platforms in the automation of the work of the department, namely the automation of the management processes of the department. The task of automation is the electronic recording of the individual plans of teachers of a higher educational institution is a key point in the analysis of the work of the entire higher educational institution as a whole. Automation of the process of fulfilling the teaching load by the teacher and filling out an individual plan: will allow you to calculate the theoretical and practical number of hours worked by the teacher automatically; will reduce the chance of error; control the publication activity of teachers of the department.

This paper presents methods for calculating the direct light transmittance of a double-grating optical filter. Based on these methods, numerical simulations were carried out, the results of which were confirmed by experiments under direct collimated white light without the influence of diffuse light using two grating smart window models with different sets of strip widths, as well as a model of a conventional window with clean glass. The results showed that by pre-calculating the angle of inclination, the widths of the strips of both gratings and their relative position, taking into account the coordinates of the building and the azimuth of the window, the filter provides angular selectivity of transmission without the use of shading/redirection devices. Window transmission with such a filter is minimized during specified times of the selected day when the need for sun protection is greatest. Due to the angular selective light transmission and the ability to adjust the inclination of the gratings to the trajectory of the Sun, such filters are promising for use in smart windows for optimal filtration of direct solar radiation indoors.

The development of modern agro-industrial complex (AIC) directly depends on the efficiency of technical solutions, especially in agricultural engineering. In the course of seasonal field work, agricultural machinery and equipment are subjected to a significant impact of various climatic factors, including solar radiation. The intensity of this impact affects the rate of wear of materials used in machinery and equipment. Especially sensitive to this impact are flexible units, which must elastically deform under compression and tension, such units become brittle and brittle. In this regard, the purpose of this work is to study and determine the effects of solar radiation on rubber seals. Object. The object of research is samples of rubber products used in agrarian engineering as a door sealant. Materials and Methods. Within the framework of this study the materials were tested on the accelerated aging unit with ultraviolet (hereinafter referred to as UV) radiation sources. Using the testing universal electromechanical machine REM5.101.0.PS the mechanical properties of the samples were analyzed. Results and conclusions. After irradiation of the sealing rubber sample for 200 hours and radiation intensity of 25 W/m² (which is ~100 times higher than the average daily power of spectra A and B of ultraviolet radiation), the reduction of tensile strength of the objects of the study amounted to 11.5 %. Local spots of different color appeared on the samples, and also based on the presented data, due to the destruction of bonds between polymer chains and the formation of cracks and microcracks in the structure of the rubber, it led to a decrease in its strength.

The work is devoted to the study of thermal processes in the “metal film-semiconductor substrate” structure when operating in a pulsed current mode. Experimental oscillograms, using the example of the Al-Si and Al-SiO₂-Si structure, show that when a series of successive rectangular current pulses are passed through the metallized area, heat accumulates in the structure, which subsequently leads to melting of the metallization system, as well as the formation of an interphase boundary Al-Si. A study of the morphology of the silicon substrate was carried out. Based on the SEM results, it was revealed that when a single current pulse with a density j=5.10¹⁰ A/m² and a duration of 500 μs was passed through, microcracks and chips were detected in silicon. Based on the temperature distribution on the surface of the semiconductor, the magnitude of mechanical stresses for the pulsed operating mode was assessed.

The work carried out an experimental study of the features of thermal destruction of Al-Ti-Si, Al-SiO₂-Si and Al-Si systems under pulsed current exposure. The dynamics of heating systems at current densities j>3•10¹⁰ A/m² and durations up to 500 μs are presented. The initial stages of melting processes in thin-film systems and the mechanisms of destruction of structures during the passage of a single rectangular current pulse have been identified. Thus, the presence of a Ti sublayer increases the thermal stability of a multilayer system; the main mechanisms of destruction of structures are associated with the phenomenon of electrical transfer, as well as the migration of molten zones in the temperature gradient field after turning off the pulse near a local heat source. The presence of a dielectric sublayer of SiO₂ leads to increased thermal loads on the metallization system (at j=3.10¹⁰ A/m² and pulse duration up to 450 μs) compared to the Al-Si structure.

The work considers small oscillations of rectilinear screw dislocation in a real crystal. Influence of the Peierls relief and short-range interaction of the screw dislocation with chains of point defects is taken into account. Using generalized susceptibility, an expression for internal friction caused by radiation friction of screw dislocation oscillations is obtained. For specific crystal, frequency dependence graphs for internal friction are plotted.

The article proposes mathematical and simulation models for studying mechanical synchronization processes that occur in asynchronous electric motors with unbalanced rotors, as well as models for studying power losses and thermal processes. As a result, dependencies were obtained that describe the relationship between power losses in the windings of vibration exciter motors, changes in rotor slip and the steady-state temperature of the motor on the one hand, and the eccentricity radius and the rigidity of the oscillating base on the other hand. As the rotor eccentricity radius increased, the phenomenon of frequency locking and the Sommerfeld effect were observed. It has been established that in these modes the total power losses can exceed the losses in the nominal mode by 2-3 times, and the current in the stator winding by 1.5-2.5 times. Experimental studies of the synchronization of oscillations of two engines on a single basis confirmed the good accuracy of the proposed models.

The article relates to the issues of optical physics. It is dedicated to the optical destruction of crystals. The main intrinsic mechanism of optical destruction is the mechanism of shock ionization or "electron avalanche". The process of destruction during shock ionization is similar to the electrical breakdown of a dielectric, when free electrons present in a small amount in the dielectric are accelerated by the electronic field of a light wave to speeds sufficient for shock ionization, giving rise to new electrons, which, in turn, accelerate, giving rise to an electronic avalanche. The peculiarity of a breakdown in an optical beam is that it is a breakdown in an alternating high-frequency field. The article also discusses some generalizing criteria that determine the applicability of a crystal for laser technology, and also provides some tips on improving the optical strength of crystalline elements.

In this work, contribution of small bending vibrations of infinite screw dislocation to internal friction is studied. Influence of the Peierls barrier and elastic interaction with point defects located along the screw dislocation line is considered. Expression is obtained for generalized susceptibility of the screw dislocation in dissipative crystal, taking into account the considered influences. Formulas are found for natural frequency of short-wave bending vibrations of the screw dislocation and contribution of bending vibrations to dislocation internal friction. Using specific crystal as example, computational experiment was carried out and graphs of frequency dependence of internal friction were constructed.

The software, implementing the mathematical model of a horse rider with four links, is presented in the article. The rider is on the horse, which is a mobile base. Unlike to the previously compiled models, this research considers the start-up operations and the accelerated motion of the horse. The local systems of coordinates are used in the software for describing dynamics of the rider on the mobile horse. These systems are more suitable for specifying the angles between the links, which corresponds to the rider’s biomechanics, whose musculoskeletal system changes these angles, applying internal efforts. The algorithm of compiling the system of equations of motion using the Lagrange equations of the second kind is implemented in the software. The programmed motion control is applied to the model. The accelerated motion of the mobile base in the form of horse as well as the angles between the rider’s links are specified by analytical differentiable functions, synthesizing the mechanical system motion resembling that of a real horse rider. Using the created software, the inverse dynamics problem has been solved. The control torques, changing the angles between the links, and the reactions of mobile base are found. The presented software, the mathematical model of the system, and the results of the research can be used for formulating recommendations for designing programming exercise machines with various working modes, which allow for rider training regardless of environment and weather conditions, horse fatigue, stable remoteness, and other aspects. It will allow horsemen to practice complex skills without fear of possible injury.

The software, implementing mathematical models of 2D and 3D active exoskeletons with electric drives, has been developed in the environment of universal computer mathematical system "Wolfram Mathematica". Each model has the same structure and includes five mobile links. Generalization of these models for the case of arbitrary finite number of n links is made. The angles, calculated between the links of the models, are used in the software described in the article, which differ these models from the existing ones. Cylindrical hinges are used for the articulated interconnection of the links in the 2D model. As for the 3D model, the spherical hinges perfectly implement all mobility degrees available in the human musculoskeletal system. This fact also differ the proposed 3D model from the existing ones, which have limited mobility and use not only spherical, but also cylindrical hinges to join some links in order to reduce dimension of the system of differential equations of motion. The comparative analysis of the obtained mathematical models for the two mechanisms in the form of Lagrange equations of the second kind is made. The system of electromechanical equations for the model using asynchronous AC motor is listed. The solution of direct and inverse dynamics problems is given for 2D model. As for the 3D model, no numerical solution of the Cauchy problem has been obtained due to large dimension of the system of differential equations of motion and limited computing power. Therefore, the idea about exoskeleton motion control model based on artificial intelligence that needs to be developed is suggested.

A fairly detailed exoskeleton variable-length link model with adjustable stiffness is considered in the article. The refinement of the previously created models in terms of discretization is made. The geometrical dimensions of the top and bottom rods, the cylinder, the cylinder covers, the piston that moves inside the cylinder with magneto-rheological fluid, as well as their masses and moments of inertia relative to coordinate axes are refined. The payload mass at the end of the link, resulting under the effect of other links or the attached actuator, is simulated by a homogeneous ball. The change of the mass in the bottom and top parts of the cylinder, and the change of the moment of inertia of the link as the magneto-rheological fluid flows inside the cylinder from one part to another through the piston with microchannels are made more precise. All these made it possible to bring the model closer to the real technical device. The mathematical model of a more accurate link model is compiled. The possible model of actuator, functioning based on the proposed refined model of the exoskeleton link with magnetorheological fluid, is described. The possibility of its functioning and the options to control it are studied. The accuracy estimation for the refined model in comparison to the previously created models is made. All the results presented were obtained in a program created in the universal software environment “Wolfram Mathematica”. The links of the proposed model can be used in designing next-generation exoskeletons, which are more comfortable comparing to the existing models.

The article describes software developed for simulating mechanical model of a training machine. The training machine is an exosuit with mobile fixing points of feet in stirrups. The exosuit includes four absolutely rigid links to which the feet, the shins, the hips, and the corps of the rider are attached. The foot attachment point corresponds to the horse’s center of mass with which the rider interacts by standing in stirrups. The motion of this point is implemented by a telescopic link firmly fixed to a supporting surface. Cylindrical hinges implement rotation of the links. It is assumed that friction in the hinges is negligible. A torque, applied to the hinge that is fixed to the supporting surface, and a lengthwise force, applied to the variable-length link section and thus changing the distance between the link ends, control the telescopic variable-length link that simulate the motion of the horse corpse. The method of programmed motion control is used in the created mathematical model. The developed software is used for synthesizing the exosuit motions close to those of a real horse rider by specifying periodic differentiable functions implementing anthropomorphic motions that subsequently can be used for inverse dynamics problem solution. The exosuit simulation results, obtained using the developed software, can be recommended for designing programmed training machines. These training machines have broad application perspectives not only in sports, but also in medical rehabilitation centers, where they can be used to restore motor capabilities of human musculoskeletal system.