Home – Home – Jornals – Avtometriya 2026 number 2

The article discusses methods for compressing information about the boundaries of areas generated by superpixel image segmentation. The representation of superpixel boundaries in the form of a four-digit image generated by boundary elements is proposed. A new compression method for this image is developed, based on constructing the trajectories of the boundary elements and then encoding them. An experimental study of the new method demonstrates its advantages over known methods of statistical coding and predictive coding, especially for large superpixel sizes.

A method of unmanned aerial vehicle trajectory planning has been developed. This method is based on the analysis of a video stream from a camera mounted on the vehicle, without using information from global navigation systems. The trajectory is planned using the determined parameters of an affine model of mutual spatial misalignment of adjacent video frames. The method utilizes the Shi-Tomassi singularity detection method, the Lucas-Kanade optical flow method, and the Kalman filter. The results of the method implementation on single-board computers and testing it under limited computing resources on real video sequences are presented.

The work is dedicated to the study of the formation and decoding of interference patterns to retrieve the phase of a probe wave. Algorithms for phase retrieval using plane and conical reference beams are considered. The paper also presents methods for improving the quality of the phase image when using a plane wave as a referent one.

The problem of effectively using domain-specific knowledge based on representative statistics is relevant from various perspectives. The methods used to utilize reliable a priori information are diverse and often specialized for each specific data analysis problem. The paper describes an approach to this problem in which solutions are developed from a unified perspective based on the use of a multidimensional analogue of the Hough transform. The approach offers new possibilities for solving two important problems: it proposes a unified method for constructing models of interactions between known particular regularities, and uses the minima of the discrepancy measure found in the Hough space as a tool for model optimization.

We consider the issues of creating algorithmic support for supervised classification problem which is the one of the central tasks of machine learning. Original procedures of logical analysis and classification of integer data represented as a set of elements of Cartesian product of finite partially ordered sets (product of partial orders) are constructed and investigated. At the training stage of the proposed procedures, the search for so-called regular representative elementary classifiers (special fragments in feature descriptions of precedents that distinguish objects belonging to different classes) is performed. An asymptotically optimal algorithm for enumerating the required elementary classifiers over a product of antichains is constructed and the results of its testing on real-world tasks are presented. Theoretical and experimental justifications for the efficiency of the new classification procedures are provided for the case when linear orders on sets of feature values are defined. The theoretical conclusions are based on the study of the metric (quantitative) properties of the set of regular representative elementary classifiers.

This paper presents the development of a neural network for predicting airfoil aerodynamic coefficients for use in the isolated-section method for calculating aircraft propeller blade characteristics. A key feature of this neural network is its ability to predict the lift and drag coefficients as functions of the angle of attack and flow Reynolds number in the form of two-dimensional raster images of pixel color distribution. This study aims to accelerate the calculation of airfoil aerodynamic coefficients while maintaining computational accuracy by replacing numerical models with a neural network. This study presents a method for airfoil parameterization, database development, and neural network architecture. The neural network training results and its ability to predict aerodynamic characteristics are presented.

The article describes the development and implementation of a microservice architecture upon which the functionality of an Intelligent Digital Twin of Plants (IDTP) is built. This system is designed for planning and simulation plant growth stages and forecasting yields, synchronized with the states of actual crops, based on ontologies and multi-agent technologies. The functional capabilities of the Manager, Ontology Constructor and Knowledge Base (OC and KB), Planning (multi-agent planner) services are examined, as well as the organization of interfaces for the interaction of the IDTP with external services and users. Examples are provided to represent the results of modeling the comprehensive state of a crop, based on real-time environmental data and forecasting the crop's condition at all stages of plant development up to harvest. The developed microservice architecture is open for integrating the IDTP with external services, such as environmental data collection, as well as recommendation systems and Agro-IoT platforms. The results of the IDTP prototype development have been passed on to several farms for experimental testing in different climatic conditions. The article discusses directions for improving the accuracy of IDTP forecasts and the potential for creating a decision support system for agronomic applications in precision agriculture based on the IDTP.

A software module for calculating the productivity and efficiency of discrete manufacturing has been developed. The module's business logic is implemented in the AnyLogic package, in which a simulation model was created and exported to a set of JAR files. The model was built for a production cell located in the IPIT-216 unit of Samara University. The simulation error was less than 15%. Unlike common commercial simulation software packages (Plant Simulation, FlexSim, Arena), the developed module enables the calculation of production KPIs in a simpler and more accessible interface, eliminating the need to purchase specialized software and undergo extensive training.

The paper considers the problem of estimating the state of charge (SoC) of lithium-ion batteries. The second-order Thevenin model based on the substitution scheme is used as a mathematical model of the battery. An extended information Kalman filter is used to estimate the state of charge of the battery

This article describes the results of a comparative analysis of classical and modified genetic algorithms for solving the problem the optimal runner position in a thermoplastic injection mold. Based on the comparison, it was found that the modified genetic algorithm converges 25% faster than the classical algorithm and allows for finding the minimum with 55% fewer calls to the criterion function, compared to the classical implementation of the genetic optimization algorithm. The issue of selecting the best criterion for solving the runner position optimization problem on a discrete finite element computational mesh is also examined.

This paper presents a method for designing support structures for 3 D printing based on topological optimization and the method of equivalent static loads. Information exchange between the optimization and numerical simulation modules of SLM process is organized for the computational process. The proposed support synthesis method effectively reduces the maximum shrinkage deflection from 0.731 mm to 0.630 mm, a reduction of 13.8%.

This article examines the dispersion coefficient of low-frequency electromagnetic waves known as whistling atmospherics or whistlers. A mathematical model describing the law of whistler frequency variation over time is proposed. An algorithm for calculating the whistler dispersion coefficient based on spectrogram analysis is also developed. This algorithm involves identifying the whistle trace using threshold filtering and clustering. The whistle trail is then straightened by transforming the coordinates, and a linear regression equation is constructed, where the cotangent of the slope provides an estimate of the dispersion. The algorithms are implemented in the ADWRK 1.0 software package using the Python programming language. The results obtained using this software package are compared with those obtained using the proposed mathematical model. It is shown that the calculation of the whistler dispersion coefficient can be more accurate using a combined method for extracting the whistler trail.

In this paper, we propose a method using conformal meshes to solve bimaterial topological optimization problems in linear and nonlinear formulations. The features of using conformal meshes in the case of solving contact problems are considered. The solution speed and results of bimaterial topological optimization on uniform and conformal meshes were compared. The results demonstrate that the use of conformal meshes increases the speed of solution of topological optimization problem by more than 12 times.

In this paper, the formation of perfect optical vortex beams is investigated based on numerical simulation for different parameters of the optical system, such as the axicon angle and the vortex beam order. The formation of beams during propagation in free space and in the focal plane is considered, which can be useful for optical capture and movement of microparticles.

This paper presents the results of developing a generator of chemically active oxygen species for air disinfection and purification for agricultural applications. The main functional element of the generator is a new highly efficient oxide photocatalyst of the Cu/ZnO-ZnAl₂O₄ system. The use of this photocatalyst ensures intensive generation of chemically active singlet oxygen, purification and disinfection of ambient air, and preservation of the high quality and freshness of agricultural products.