Information Technologies and Intelligent Decision Making Systems

The purpose of the work is to create an intelligent technology for obtaining oxide coatings with specified properties through an automated system being developed. The developed intelligent system implements the method of micro-arc oxidation to obtain protective oxide coatings on products made of aluminum, titanium or their alloys. The intelligent system consists of hardware, software, and information content. The software is developed in the LabVIEW graphical programming environment. The intelligent application contains three subroutines: identification of the electrophysical model parameters and optimization of process parameters to obtain the required properties of oxide coatings; visualization of the coating parameters dependences in real time on influencing factors. The algorithm of the developed automated system functioning is presented. The presence of an intelligent application allows feedback for software, in which the process current source can change the oxidation mode depending on the coating state at a given time, taking into account the required target coating parameters (thickness, porosity, hardness, etc.). The advantage of the proposed intelligent system is the possibility of implementing a regime of controlled synthesis of oxide coatings with the required properties. In turn, it makes it possible to increase the reproducibility of the MАO coatings parameters, reduce the time for testing the technological process.

This paper describes the process and results of developing a methodology that contributes to the full implementation of an object storage file management system. The methodology is designed for use in the implementation of information systems with a microservice architecture. The sites of the digital platform became the pilot site of the study. A practical request for the development of the methodology is the need for digital platforms to organize the storage of website content and organize access to it. As a result of the conducted research, the effectiveness of the methodology for optimizing the process of filling in the content on digital platform sites when using a file management system is substantiated. The application of the file management system implementation methodology is designed to optimize the work with content. The solution can be scaling for other systems in the context of micro service architecture. This will require partial adaptation of the methodology to the current requirements and needs of the systems. The fact that the implementation of the methodology reduces the time for the implementation of the content filling process allows us to assume that the methodology can be tested in other systems after the analysis. The developed methodology has a great potential for development due to the expansion of functionality.

The work examines time series clustering using various machine learning approaches. The purpose of the study is to compare the performance of K-shape, K-means, and hierarchical density-based spatial clustering with noise (HDBSCAN) algorithms. Clusters were measured using the Rand Index, Adjusted Rand Index, Adjusted Mutual Information (AMI), and measures based on the electrocardiogram (ECG), ArrowHead, and SharePriceIncrease datasets. Noise was added to the time series data and clustering was performed on the noisy data. The resulting clusters were compared to the original clusters using Bcubed metrics to evaluate the robustness and accuracy of clustering algorithms in noise engineering. The results of the study will shed light on the effectiveness of these clustering algorithms in detecting anomalies in time series data. Additionally, the influence of colored noise on the accuracy and stability of the clustering algorithm will be determined.

This article presents an open platform tailored for the analysis of transportation data. It provides an overview of the primary types of data involved, including individual transport mobility data and aggregated data organized by time intervals. The platform adopts a modular architecture. The first module is dedicated to data loading, storage, and sampling for practical problem-solving. Subsequently, the second module undertakes semantic data validation, ensuring the integrity of subsequent modules by identifying and rectifying errors encountered during data collection. The third, analytical, module allows experts to apply data analysis methods to solve problems of urban infrastructure management. A methodology grounded in the characterization of normality and the detection of anomalies is proposed for problem resolution. Furthermore, the fourth module specializes in visualizing analysis outcomes through diverse formats such as tables, graphs, and maps. The fifth module, an API component, extends the platform's accessibility and permits the integration of new functionalities into the analytical module, along with novel data management techniques. The operation of the platform is demonstrated by the example of searching for anomalies in traffic flows, clustering city districts, as well as visualizing the volume of passenger traffic at train stations in the Moscow region.

The transformations of signals and the properties of the radiation beampattern in a frequency-diversity antenna array are investigated. The effect of focusing and steering on the beampattern is being investigated. A conventional uniform coherent FDA antenna with a linear frequency plan is considered in detail, and its directional patterns with focus only on transmission, only on reception, and full focus are studied. FDA antennas with nonlinear and symmetrical frequency distributions over the elements are considered. The effect of matched filtering on the appearance of the radiation pattern has been studied.

The paper considers the problem of improving the cruise control system used to control the speed of a truck. It is proposed to supplement the existing control system with a supervisor that adjusts the parameters of the classic PID controller in real time depending on the current state and driving modes of the vehicle. To ensure effective work of the supervisor in conditions of incompleteness and inaccuracy of incoming information, it is proposed to use fuzzy logic methods. Three most frequently used variants of the fuzzy supervisor implementation are considered: the PID controller coefficients are calculated by the supervisor based on the error signal, the PID controller coefficients are adjusted by the supervisor based on the error signal, the PID controller coefficients are calculated by the supervisor based on the difference between the output of the control system and the reference model. The parameters of the PID controller and the fuzzy supervisor are adjusted to ensure the minimum value of the transition process time. For each of the considered variants, the results of computer modeling of the processes of changing the speed of a truck are presented. The most effective algorithm for controlling the speed of a truck is proposed, as well as directions for further improvement of the cruise control system.

The paper describes the usage of blockchain technology to protect paper documents from forgery surpassing the characteristics of existing protection methods. Modern methods of paper documents protection and blockchain specifics regarding to this purpose are observed. A new approach to verifying the authenticity of paper documents using blockchain technology is proposed and a software tool that implements this approach is developed. Results of the tool application and its advantages in compare to existing methods are discussed.

The essence of the dimensionality reduction process is the transition to a more concise set of indicators in such a way that the associated loss of information present in the source data is minimized. This article proposes methods and algorithms for reducing the dimension of the original feature description space for the problem of pattern recognition. Four types of applied problems of reducing the dimension of the analyzed feature space are defined. The mathematical model underlying the construction of one or another dimension reduction method has been determined. For each method discussed below, a step-by-step algorithm is given that allows you to quickly and efficiently translate these methods into programming languages for various types of computers.

The article discusses the necessity of automating the process for checking and evaluating students’ written work in educational institutions. It proposes the recognition of handwritten texts through computer vision methods, specifically by considering the application of convolutional neural networks (CNNs). An architecture for a CNN capable of recognizing letters from both the English and Russian alphabets is developed. The network is trained on extended datasets to enhance recognition accuracy and prevent overfitting. The authors have designed a conceptual model for an information system that processes students’ tests via an adaptive website. This includes various user scenarios, context diagrams for uploading test answers, and procedures for text recognition in images. The article also deconstructs the contextual structure of the written work assessment process. Automating the checking of written work promises to expedite and simplify the assessment process for students’ assignments.

The paper is devoted to visualization of continuous integration pipelines of the Jenkins system. When working with the Jenkins system additional tools are needed to automate the process by visualizing pipelines. There are a number of existing Jenkins plugins for pipelines visualization, however all of them have specific shortcomings. Based on the analysis of existing solutions requirements for a new plugin were formulated. Also presented in the paper are the architecture and implementation details of the developed plugin which allows to visualize pipelines both in the form of a graph and a Gantt chart as well as provides the user with metadata, crash and restart information. Results of the plugin integration into Jenkins pipeline prove its effectiveness due to reduction of efforts on integration pipeline analysis.

An algorithm is proposed for eliminating refractive distortions caused by the oscillating surface of a liquid when studying the brain of a live mouse. Studies like this in mice allow monitoring neuronal activity in a living organism, and the fluid is needed to ensure that the brain remains in its natural environment. However, the presence of fluid flow causes distortions that significantly complicate tracking waves of neuronal activity. The goal of the present work is to remove effects that displace and distort images of individual parts of the brain, and, in fact, bring the entire image to a static picture. The proposed algorithm, based on tracking individual parts of images, gives a 10% improvement in approximation to a static picture compared to the original recording.

This paper presents an approach to apply quantum Fourier transform (QFT) to image processing using quantum computing. The use of quantum computing for image analysis and processing is becoming increasingly relevant in modern science and technology. A quantum QFT circuit is presented, implemented using the Qiskit framework, which is a tool for programming quantum computers. The paper presents the basic steps of QFT and their application to a state vector representing the pixel intensities of an image. The influence of quantum transformation on the image structure is studied and the results are presented in the form of graphs and visualizations. In addition, we have introduced QFT quantum circuit inference capabilities for a more visual representation of the algorithm. The results highlight the potential of quantum computing in the field of image processing and open new prospects for the use of quantum technologies in the field of computer vision.

The article formulates and solves the problem of choosing a set of countermeasures to protect the software of a corporate information network. The problem is formalized by analogy with the open model of the assignment problem in the integer formulation. An example of a numerical implementation of the problem of choosing the optimal software protection mechanism based on the decay vector method is considered. The implementation of the algorithm of the method is written in the Java language. The resulting solution is a set of countermeasures recommended to neutralize security threats to the software.

The article explores the use of neural networks to solve the problem of determining boundaries of meshes of perineuronal nets. The confocal stacks’ image layers of rat brains are used as initial data. This article presents a comparison of two alternative methods to solving the problem. The first method is based on the generation of boundaries through the use of a neural network based on the DCGAN architecture. The second method is based on solving the problem of semantic segmentation using a neural network based on the U-Net architecture, that is widely used in biomedicine. For both neural networks, architectural changes are presented to achieve greater generalizability of the models. Some learning strategies are considered to solve the problem of overfitting that is typical for small samples. Both solutions showed results comparable in quality to the semi-automatic algorithm. A solution based on the U-Net architecture provides good tools for further solving the problem of boundary ambiguity and allows customizing the algorithm for various criteria for boundaries of the perineuronal nets.

This study is devoted to the development of improved methods for diagnosing cattle diseases based on the principles of neutrosophic sets and Sugeno fuzzy inference. The study proposed new algorithms and models that can effectively process fuzzy and uncertain information characteristic of veterinary diagnostics. The goal of the work is to create a diagnostic system that will have high accuracy and the ability to adapt to various conditions and characteristics of specific disease cases. The expected result of the study is aimed at developing an effective tool for the early detection of diseases in cattle, which will significantly improve the efficiency of veterinary practice and animal welfare.

The article discusses the technique of parametric adaptation of big data. The basis of this technique is vector rank analysis, developed within the framework of the technocenological theory of Professor B.I. Kudrin. The novelty of the technique lies in the fact that for the first time, together with technocenological methods, the possibility of using elements of artificial intelligence has been realized. The main stages of the methodology are: loading “raw” data, forming matrices of verified and approximated data, as well as checking their adequacy based on vector rank analysis methods. These steps are performed in parallel using ChatGPT based on the proposed artificial intelligence methods.

Achieving energy and resource efficiency through the application of modern information technologies, calculation algorithms, automatic management and control systems to the process of waste gas production from solid household waste is an urgent issue today. In the article, the author analyzed the system of automatic management and control of the process of heliothermic processing of solid household waste based on information technology. The thermal scheme of the automatic and control system of the exhaust gas production process from municipal solid waste is proposed. ATmega8 microcontroller, which allows automatic management and control of all technological and thermochemical processes of waste gas production, was used in the heliothermic waste gas processing unit. At the same time, all technological processes of waste gas production, collection and use are automated, anaerobic fermentation (50 ÷ 55 ℃) in the device, automatic control of heat distribution processes between layers in the waste reactor, waste protection, technological process parameters, order charging and energy saving is ensured. Automation and control of the solid waste heliothermic processing plant is carried out by the application of information technologies, the information resource of the system, working pressure, temperature, temperature range, humidity, consumption of raw materials, the level of the mixture in the waste reactor, the pH indicator, as well as the control of the operation of the valves.

The article discusses the main tasks of machine learning. The functional structure of a computer algorithm for solving machine learning problems and a data mining model are considered. The solution of the simplest machine learning problem using the classification method of linear regression is proposed. The analysis of existing learning algorithms based on a decision tree is carried out. Based on the analysis performed, a decision tree was selected for implementation using the so-called C4.5 algorithm. The article builds a decision tree using specific training data. The application of a simple and understandable algorithm for building trees is to create all possible trees, calculate the number of erroneously classified data for each of them and select a tree with a minimum number of errors. As a result, an optimal learning algorithm is formed for the decision tree in terms of data errors in the learning process. The top-down algorithm implemented in the article for building a decision tree selects the attribute with the largest increase in information at each step. Entropy is used as a metric of the amount of information in the training data set D. In the process of implementation, this algorithm is analyzed to identify opportunities for its application to a more complex task.