Arai Tolegenova

In modern electric power systems, accurately identifying short-circuit conditions is essential for reliable network monitoring and protection. This study proposes a hybrid machine learning framework for multi-class fault mode classification in three-phase electric networks. The experimental dataset included 15 operating modes, including normal mode, single-phase mode, phase-to-phase mode, three-phase mode, and ground fault mode. Input variables included phase current measurements and positive, negative, and zero-sequence current components. The preprocessing pipeline included logarithmic transformation, standardization, and correlation-based feature selection. The processed data was split into training and independent test subsets in an 80/20 ratio. Next, k-means clustering with 15 clusters was used to generate additional structural features, while random forest and extreme gradient boosting (XGBoost) algorithms created probabilistic class predictions, which were integrated by a logistic regression metaclassifier. According to the adopted testing protocol, the proposed hybrid model achieved an accuracy of 96.98%, a macro-precision of 0.973, a macro-recall of 0.969, and a macro-F1-score of 0.968, outperforming the baseline random forest model, which achieved an accuracy of 93.91% and a macro-F1-score of 0.937. These results demonstrate the potential of the proposed multi-level framework for accurate multi-class fault diagnosis in three-phase electrical networks.

The object of this study is text-based news content distributed through online media and social media platforms, presented as vector objects formed on the basis of unstructured text data and used for subsequent automated analysis. The problem solved in this study is the limited effectiveness, stability, and generalizing ability of traditional machine learning methods in detecting fake news, especially in conditions of heterogeneous datasets, noisy textual characteristics, and dynamically changing linguistic patterns, which negatively affects the quality of classification. The article proposes a method for improving the efficiency of machine learning based on the combined use of SVM and the AdaBoost algorithm. To form an informative representation of text data, complex preprocessing and feature extraction using the TF-IDF model are used. The experimental verification of the method was performed on four open datasets: ISOT, Kaggle, News Trends and Reuters. The results show that the proposed SVM ensemble model with AdaBoost is superior to the basic SVM classifier and a number of traditional algorithms. Accuracy increased from 0.8175 for the base model to 0.83 for SVM+AdaBoost, while memorization increased by 4.02%, average memorization accuracy increased by 2.22%, and the F1 index increased by 1.84%, while the stability of the test accuracy decreased only slightly by 0.19%. The improvement is explained by AdaBoost's adaptive enhancement of the contribution of hard-to-classify objects and a reduction in the number of errors with moderate computational complexity. The developed approach can be effectively applied in automated systems for monitoring news content and social networks in the presence of marked-up text data and limited computing resources

This research focuses on the development of a microstrip patch antenna (Ant8) for integration with Intelligent Reflective Surfaces (IRS), aiming to enhance wireless communication systems, especially in dynamic RF environments. The main problem addressed is the limited bandwidth and directional control in current compact antennas for IRS-assisted systems. To solve this, a novel dual-band antenna was proposed and optimized, incorporating vertical VIAs and Minkowski fractal slots to expand bandwidth and achieve circular polarization. Numerical simulations and experimental measurements confirmed stable operation in the 2.22–2.36 GHz and 2.88–3.81 GHz bands, with enhanced signal gain (up to 6.2 dBi), improved return loss, and narrow directional patterns. These results are explained by the antenna’s unique structural features: vertical VIAs induce a lower resonance, while fractal slots increase current path length without enlarging physical dimensions, enhancing bandwidth. Key distinguishing features – dual-band response, circular polarization, and compact design – allowed effective IRS integration, enabling dynamic beam steering and interference mitigation through phase control. The results are especially effective in multipath and shadowed environments, where signal reliability is critical. Practical applications include telemedicine (ensuring high-speed and stable data transfer), smart buildings (enhancing coverage and reducing interference), and autonomous transportation (reliable vehicle-infrastructure links). The antenna – IRS system is well-suited for mobile platforms, urban deployments, and IoT networks within emerging 5G and 6G infrastructures.

This research focuses on optimizing ultra-wideband (UWB) antennas, which are critical in modern communication systems due to their wide frequency range (3.1–10.6 GHz) and high data transmission capabilities. The study addresses the challenge of optimizing key antenna parameters – such as return loss, peak gain, and radiation efficiency – while also ensuring energy efficiency and network longevity. Traditional optimization methods, such as LEACH-C, often fail to balance these factors, leading to suboptimal performance. To solve this problem, the researchers developed the Generalized Position-based Optimization Neural Network (GPON) for UWB antenna optimization. They also evaluated the Position-based Hybrid Neural Network (PAN) method, comparing its performance with existing algorithms including LEACH-C, Firefly Algorithm (FA), HFAPSO, FA-ANN, and HWOABCA. The GPON model reduced return loss to 25.5 dB at 3.5 GHz and improved peak gain to 4.2 dB i, while maintaining 92 % radiation efficiency. In contrast, PAN demonstrated a 15–25 % improvement in residual energy and extended network lifetime by 20 % compared to LEACH-C. These improvements were due to the integration of advanced neural network techniques in GPON and the effective use of positional data in PAN, enabling more precise and adaptive optimization. The ability to balance multiple performance metrics simultaneously – a challenge previous models struggled with – is a key feature. This balance is crucial for UWB antennas in communication systems where both performance and energy efficiency are vital. The findings are especially relevant for practical applications in wireless sensor networks, mobile communications, and radar systems, requiring long-term network reliability and optimal antenna performance

This article studies the influence of solid-phase (type 1 samples) and melt-quenching (type 2 samples) technological modes of obtaining Na₃Fe₂(PO₄)₃ polycrystals on their structures and ion-conducting properties. α-Na₃Fe₂(PO₄)₃ polycrystals of the 1st type are formed predominantly under an isothermal firing regime, and the synthesis of the 2nd type is carried out under sharp temperature gradient conditions, contributing to the formation of glassy precursors possessing a reactive and deformed structure, in which the crystallization of crystallites occurs faster than in precursors obtained under isothermal firing. The elemental composition of α-Na₃Fe₂(PO₄)₃ type 2 polycrystals is maintained within the normal range despite the sharp non-equilibrium thermodynamic conditions of synthesis. The microstructure of the type 1 Na₃Fe₂(PO₄)₃ polycrystals is dominated by chaotically arranged crystallites of medium (7–10 μm) and large (15–35 μm) sizes, while the polycrystals of type 2 are characterized by the preferential formation of small (3–4 μm) and medium (7–10 μm) crystallites, causing uniaxial deformations in their structure, which contribute to a partial increase in their symmetry. The advantage of type 2 polycrystals is that they have higher density and conductivity and are synthesized faster than type 1 samples by a factor of 4. The article also considers the issues of crystallization in a solid-phase precursor from the classical point of view, i.e., the process of the formation of small solid-phase nuclei in the metastable phase and their growth to large particles due to association with small crystallites using phase transitions. Possible variants and models of crystallite growth in Na₃Fe₂(PO₄)₃ polycrystals, as well as distinctive features of crystallization between two types of samples, are discussed.

The object of this study is a new design of an overhead microstrip antenna, including Epsilon Negative (ENG) metamaterials and L-shaped parasitic overlays aimed at achieving dual-band operation and improving the performance of modern wireless communication systems. The research is aimed at solving the problem of limited bandwidth and dual-band functionality in conventional antenna designs. This study solves the problem of limited bandwidth and dual-band functionality in conventional microstrip patch antenna designs, which are crucial for modern wireless communication systems. The inclusion of L-shaped parasitic overlays effectively expands the bandwidth in each frequency range. The results indicate significant improvements: at the upper resonant frequency, a 10 dB return loss bandwidth of 27.84 % (2.88–3.81 GHz) and a 3 dB axial ratio bandwidth of 5.05 % (2.90–3.05 GHz) are achieved, while at the lower resonant frequency, a return loss bandwidth is 10 dB, which is 6.11 % (2.22–2.36 GHz). These improvements indicate a significant increase in antenna performance compared to conventional designs, which is confirmed by comparing the simulation results with the measurement results. Distinctive features contributing to these results include the use of ENG metamaterials, which improve electromagnetic properties and signal propagation; vertical transitions, which provide efficient dual-band operation; and L-shaped parasitic overlays, which significantly expand bandwidth. These characteristics combine to eliminate the limitations of traditional designs, which makes the proposed antenna suitable for use in a wide frequency range in modern wireless communication systems.

In this paper, the object of the research is the implementation of artificial neural networks (ANN) on FPGA. The problem to be solved is the construction of a mathematical model used to determine the compliance of FPGA computing resources with the requirements of neural networks, depending on their type, structure, and size. The number of its LUT (Look-up table – the basic FPGA structure that performs logical operations) is considered as a computing resource of the FPGA. The search for the required mathematical model was carried out using experimental measurements of the required number of LUTs for the implementation on the FPGA of the following types of ANNs: – MLP (Multilayer Perceptron); – LSTM (Long Short-Term Memory); – CNN (Convolutional Neural Network); – SNN (Spiking Neural Network); – GAN (Generative Adversarial Network). Experimental studies were carried out on the FPGA model HAPS-80 S52, during which the required number of LUTs was measured depending on the number of layers and the number of neurons on each layer for the above types of ANNs. As a result of the research, specific types of functions depending on the required number of LUTs on the type, number of layers, and neurons for the most commonly used types of ANNs in practice were determined. A feature of the results obtained is the fact that with a sufficiently high accuracy, it was possible to determine the analytical form of the functions that describe the dependence of the required number of LUT FPGA for the implementation of various ANNs on it. According to calculations, GAN uses 17 times less LUT compared to CNN. And SNN and MLP use 80 and 14 times less LUT compared to LSTM. The results obtained can be used for practical purposes when it is necessary to make a choice of any FPGA for the implementation of an ANN of a certain type and structure on it

Nonlinear optical effects in optical waveguides play an important role in the development of fiber and integrated optics systems for optical communication and information processing. On the one hand, nonlinear effects impose restrictions on the radiation power that can be transmitted through an optical fiber or light guide. In this paper, the problem of the occurrence of the phenomenon from the stimulated Brillouin scattering (SBS) effect is investigated using two optical sources of rays in a single-mode optical fiber at joint waves of 1310 nm and 1550 nm. Due to the fact that in all trunk fiber-optic lines, the intensity and energy of input signals is limited due to the influence of SBS, methods are currently being sought to reduce the influence of this phenomenon. It was found that the energy of the input beam in the combined propagation of the compound did not reach the value of the threshold of SBS in the values of 25 dBm and 27 dBm due to the discrepancy between the experimental results and the results of the model. The SBS effect was not observed when the threshold threshold of 15 dBm and 27 dBm – SBS was reached when combining a dual beam along a single optical fiber in one direction. As a result, by double integration, the value of the SBS threshold was raised, and the direction for future scientific research was determined. If the possibility of increasing the threshold of SBS is proved, then increasing the distance of amplifiers in the main networks, respectively, its economic effect increases. In addition, it can be noted that there are no scientific papers devoted to the study of the effects of optical nonlinear effects by combining and distributing these two compounds along a single optical fiber. This article discusses the issues of improving the capacity and determination of the threshold for stimulated Brillouin scattering. To increase the power threshold is invited to consider the dependence of the phase modulation frequency of the spectral width of the laser radiation. In addition, the fact that the SBS threshold did not reach the value at an energy of 27 DB in the case of the double-beam distribution can be proved by discrepancies in comparison with the results of experimental studies and the results of the model that determines the SBS effect

Time series data analysis and forecasting tool for studying the data on the use of network traffic is very important to provide acceptable and good quality network services, including network monitoring, resource management, and threat detection. More and more, the behavior of network traffic is described by the theory of deterministic chaos. The traffic of a modern network has a complex structure, an uneven rate of packet arrival for service by network devices. Predicting network traffic is still an important task, as forecast data provide the necessary information to solve the problem of managing network flows. Numerous studies of actually measured data confirm that they are nonstationary and their structure is multicomponent. This paper presents modeling using Nonlinear Autoregression Exogenous (NARX) algorithm for predicting network traffic datasets. NARX is one of the models that can be used to demonstrate non-linear systems, especially in modeling time series datasets. In other words, they called the categories of dynamic feedback networks covering several layers of the network. An artificial neural network (ANN) was developed, trained and tested using the LM learning algorithm (Levenberg-Macwardt). The initial data for the prediction is the actual measured network traffic of the packet rate. As a result of the study of the initial data, the best value of the smallest mean-square error MSE (Mean Squared Error) was obtained with the epoch value equal to 18. As for the regression R, its output ANN values in relation to the target for training, validation and testing were 0.97743. 0.9638 and 0.94907, respectively, with an overall regression value of 0.97134, which ensures that all datasets match exactly. Experimental results (MSE, R) have proven the method's ability to accurately estimate and predict network traffic

A distinctive feature of this work is to select magnetic system parameters to stabilize nano-satellite of CubeSat format, taking into account various constraints. Analysis of the physical conditions and technical problems associated with the resource constraints of this class of spacecraft was performed. Control of rotational speed and orientation of the rotation axis is made by magnetic coils interacting with the magnetic field of the earth. Results of the simulation of the satellite motion relative to the center of mass at using the proposed control algorithm “B-dot” are demonstrated. We consider two autonomous algorithms of control by current coils. To work with orientation magnetic system using the first algorithm it is necessary to evaluate the vector of the angular velocity of the satellite. The second algorithm does not require evaluation of the angular velocity. The paper also analyzes the energy consumption of the orientation magnetic systems depending on the selection of the desired direction of the rotation axis to stabilize very small CubeSat spacecraft.