ru ro

E-Journal №3(71)2026

"PROBLEMS of the REGIONAL ENERGETICS (https://doi.org/10.52254/1857-0070.2026.3-71)"

CONTENTS

1	Neural Network-Based Stator Current Control for Rated Power Regulation of Permanent Magnet Synchronous Generator Wind Turbines with Differential Evolution Pitch Optimization Authors: Hazem Hassan Ali Sayed Energy Management Department, Luminous Energy Solutions, Calgary, Canada
	Abstract: The main objective of this study is to achieve precise regulation of the Machine Side Converter (MSC) stator quadrature-axis current in a Permanent Magnet Synchronous Generator (PMSG) wind turbine, in order to reach rated maximum power at wind speeds exceeding the rated limit while preventing excessive current surges that could damage converter components, since a comprehensive comparison between adaptive Neural Network (NN) and Proportional Integral (PI) controllers for this purpose remains insufficiently addressed in the literature. These objectives were achieved by solving the following tasks: an accurate pitch angle estimation method was developed for wind speeds exceeding the rated value using the Differential Evolution (DE) algorithm; a PI controller was tuned to maintain the PMSG electrical angular speed at its rated value through the DE-optimized pitch angles; an adaptive NN controller, trained online using online delta-rule backpropagation algorithm, was developed to directly regulate the MSC stator quadrature-axis current; and both control schemes were implemented in MATLAB/Simulink under variable wind speed and generator inertia conditions. The most important results are that the MSC-based NN controller achieved rated maximum power with zero overshoot, compared to 72.9% overshoot in active power obtained with the PI-based speed regulator, and markedly reduced overshoots in the stator current, mechanical torque, rotor speed, and DC-link voltage. The significance of the obtained results lies in showing that direct neural-network-based current regulation protects MSC components from excessive current stress, extends system lifespan, and offers a scalable solution for large-scale wind energy deployment.
	Keywords: online delta-rule backpropagation algorithm, stator quadrature-axis current, rated power regulation, evolutionary optimization, variable speed wind turbine. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.01
2	Transformation of Radiation–Convective Heat Transfer and Thermomechanical State of Industrial Steam Generators under Furnace-Space Ballasting with Carbon Dioxide from Biogas Authors: Baranyuk О., Rachуnskyі A., Drachuk O., Tуkhokhod V., Tkachenko S. National Technical University of Ukraine «Igor Sikorsky Kyiv Polytechnic Institute» Ukraine, Kyiv
	Abstract: The purpose of this study is to provide a scientific substantiation of the mechanisms governing changes in the structure of local heat absorption and the kinetics of pollutant formation when natural gas is replaced by biogenic gas mixtures ballasted with carbon dioxide. To achieve this objective, the following tasks were accomplished: numerical simulation of three-dimensional combustion processes was performed, and an accurate assessment of the thermomechanical state of critical boiler unit components was conducted. The integrated numerical approach combines Computational Fluid Dynamics (CFD) and Finite Element Analysis (FEA) methods over an extended load range (40–100% of nominal capacity). The scientific novelty of the results lies in the identification of the physical effect of radiation heat transfer intensification within the furnace volume due to an increase in the partial pressure of triatomic gases (CO₂), which compensates for the reduction in the adiabatic flame temperature of biogas. It has been theoretically demonstrated that an increase in the volumet-ric flow rate and emissivity of combustion products leads to a redistribution of thermal load along the boiler circuit, causing a stable increase in temperature within the convective pass by 20–25 °C. This effect ensures that the flue gas temperature remains above the dew point, thereby eliminating the risk of low-temperature corrosion of heating surfaces under deep load-following operating conditions. The practical significance of the study lies in establishing the regularities of local heat flux homogenization, which reduces temperature gradients and the amplitude of peak thermomechanical stresses in the metal of waterwall tubes and drums by 15–20%. The obtained results provide a scien-tific basis for predicting the service life and thermal reliability of pressure-retaining components in the context of energy sector decarbonization.
	Keywords: biogenic gas mixtures, computational fluid dynamics (CFD), thermostructural analysis (FEA), radia-tive heat transfer, deoxification, decarbonization. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.02
3	Optimization of Weakly Coupled Wireless Power Transfer Systems for Electric Vehicle Charging Stations Authors: Panteleev V.I., Sizganova E.Yu., Petukhov R.A., Kovalenko I.V. Siberian Federal University Krasnoyarsk, Russian Federation
	Abstract: The main objective of this research is to develop a methodology for selecting the optimal resonant frequency of weakly coupled wireless power transfer systems for electric vehicle charging stations, accounting for skin effect, proximity effect, and non-sinusoidal excitation. A mathematical model considering frequency dispersion of lumped parameters and spatial current density redistribution in planar spiral coils was developed. Numerical simulations were performed using finite element analysis with adaptive boundary layer meshing, followed by circuit-level co-simulation. The combined influence of frequency, coupling coefficient, and compensation topology on system efficiency was systematically analyzed. The most significant results include quantitative characterization of parameter variation in planar coils across the frequency range, revealing monotonic increase in AC resistance and decrease in self-inductance with rising frequency, while the coupling coefficient remains invariant for a fixed geometry. A lower effective frequency boundary was established based on an analytical quality criterion, ensuring high link efficiency and overall charging efficiency. The filtering capability of the series-series compensation topology was verified, demonstrating effective suppression of higher harmonics under square-wave excitation. The significance of these results lies in the development of a computational toolkit that establishes an explicit relationship between coil geometry, skin-depth dynamics, and system efficiency. This enables optimization of output characteristics without excessive shielding or artificial enhancement of magnetic coupling. The proposed methodology provides a reliable framework for designing and optimizing weakly coupled wireless power transfer systems under realistic operating conditions.
	Keywords: wireless power transfer, series-series compensation, skin effect, finite element analysis, COMSOL, MATLAB, weak coupling, electric vehicles. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.03
4	Data-Driven Zone-Level Temperature Prediction for Energy-Efficient and Demand-Responsive HVAC Control in Smart Buildings Authors: Subramanya S. A., NarasimhaIyengar N.A. B.M.S. College of Engineering, Bengaluru, India
	Abstract: The main objective of this study is to develop an accurate and scalable data-driven methodology for predicting zone-level indoor air temperature (IAT) to support sustainable HVAC operation in smart buildings. These objectives were achieved by collecting IoT-enabled time-series thermal data from multiple air handling unit-based zones within an educational building. A rule-based zone grouping strategy based on floor level, orientation, solar exposure and usage characteristics was implemented to address spatial variability. Three deep learning architectures, namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) and Recurrent Neural Network (RNN), were developed and evaluated using real operational data. The models were used to capture temporal thermal dynamics and to assess prediction accuracy and computational efficiency under both normal and demand re-sponse (DR) operating conditions. The most important results indicated that the LSTM model achieved the best prediction performance, with a Mean Absolute Error (MAE) of approximately 0.1°C and coefficient of determination (R²) values of up to 95%. The developed models also showed fast prediction times of approximately 4–5 seconds, making them suitable for real-time HVAC applications. In addition, the prediction performance remained stable during DR events, confirming the robustness of the developed models under dynamic operating conditions. The significance of the obtained results lies in showing that zone-level data-driven thermal prediction combined with intelligent zone grouping can provide reliable short-term IAT forecasting for sustainable and responsive HVAC operation in smart buildings.
	Keywords: building energy management, demand response, indoor thermal modeling, real-time prediction, time-series forecasting, zone grouping. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.04
5	Improving the Efficiency of Direct Current Contactors Operation through the Use of an Arc Extinguishing Chamber with Permanent Magnets Authors: Kirillov I.V., Dergachev P.A. National Research University «MPEI» Moscow, Russian Federation
	Abstract: The objectives of this study are to prove the operability and efficiency of an arc extinguishing chamber equipped with neodymium permanent magnets of grade N35 specifically for application in direct current circuits of railway locomotives, and to confirm that a direct current breaking arc can be extinguished solely by a non-uniform magnetic field created by these magnets. To achieve this goal, a series of calculations based on a computer model of the full breaking arc extinction process in the aforementioned chamber was carried out for currents ranging from 88 A to 0.4 kA. Several important tasks were solved during this research, and we also focused on the analysis of existing DC contactors and their arc systems, the structural design of the chamber, increasing the extinction speed, and ensuring bidirectional commutation, which is critical for traction motors operating in regenerative braking mode. The key results obtained during the simulation are successful extinction solely by the non-uniform field of the N35 magnets, an extinction time of ~5 ms at 0.4 kA, and proven operability during changes in current direction. The practical significance of this work lies in applying the chamber in railway DC contactors to optimize the overall layout of electrical apparatus and power supply systems, reduce the associated technological requirements for switching equipment production, enhance reliability and operational safety due to higher speed, and scale the proposed solution for a series of DC contactors for both power and auxiliary circuits.
	Keywords: magnetic blow-out, breaking electric arc, permanent magnets, railway contactors. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.05
6	Development of a Method for the Analytical Synthesis of Standards for Navigation Control Systems of Unmanned Aerial Vehicles with a Deficit of Informative Features Authors: 1Sotnikov A.М., 2Tiurina V.Yu., 3Lukyanova V.A., 4Zhelanov O.O., 5Antonets V.V.,5Bashkatov Ye.G. 1Kharkiv National Air Force University (KNAFU), Kharkiv, Ukraine 2National Defence University of Ukraine: Kyiv, Ukraine 3V. N. Karazin Kharkiv National University, Kharkiv, Ukraine 4Kharkiv National University of Radio Electronics, Kharkiv, Ukraine 5Educational and Scientific Institute of Ensuring State Security the National Academy of the National Guard of Ukraine, Kharkiv, Ukraine
	Abstract: The aim of this article is to synthesize reference images with high invariance and sufficient information content to improve the operational stability of unmanned aerial vehicles (UAVs) under conditions where informative features of target objects are insufficient. This goal is achieved through parametric optimization of the transformation order and regularization of the object image reconstruction process in fractional Fourier transform space. A method for analytically synthesizing reference images is proposed, enabling the extraction of the integral phase structure of an object while preserving the information invariance of the signal under critical noise and a deficiency of informative features. An analytical expression for synthesizing an optimal reference image is derived. The solution effectively suppresses noise through the use of a regularization function and a regularization coefficient, followed by the return of the reference image to the coordinate plane. An analytical model of an ensemble of optimal reference images is synthesized. It is shown that this approach ensures high noise immunity of machine vision systems under changing viewing angles and weather conditions, while maintaining the low computational complexity of image matching algorithms. The effectiveness of the developed method was assessed using noise immunity criteria and a regularized structural quality criterion for the reference standards. It was demonstrated that, under low signal-to-noise ratio conditions, the use of synthesized reference standards ensures reliable system operation, while maintaining a standard deviation of 2–4 meters. The effectiveness of the method was confirmed through numerical simulation of real scenes in the Python software environment.
	Keywords: unmanned aerial vehicle, machine vision system, reference object, informative features, destructive impact. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.06
7	Influence of Low Heat Rejection Combustion Chamber on the Behavior of Coffee Husk Methyl Ester Biodiesel in a Direct Injection Diesel Engine Authors: 1Vasanthkumar P., 2Venkatesan K., 3Shanmuganandam K., 4Parthasarathi R., 5Balu P. 1SRM Institute of Science and Technology, Ramapuram Campus, Chennai, India 2Vel Tech High Tech Dr.Ragarajan Dr.sakunthala Engineering College, Avadi, Chennai, Tamil Nadu, India 3Bharath Institute of Higher Education and Research, Chennai, India 4Ganapathy Chettiar college of Engineering and technology, Paramakudi, India 5Bharath Institute of Higher Education and Research, Chennai, Tamil Nadu, India
	Abstract: The main objectives of the study were to investigate the effect of a Low Heat Rejection (LHR) com-bustion chamber on the performance, combustion, and emission characteristics of a direct injection (DI) diesel engine fueled with Coffee Husk Methyl Ester (CHOME) biodiesel blends, and to deter-mine the optimum blend for efficient engine operation. The study addressed several key challenges associated with biodiesel-fueled diesel engines, including lower thermal efficiency, higher fuel consumption, increased nitrogen oxide (NOx) emissions, and significant heat loss through engine com-ponents. To overcome these issues, a 0.3 mm partially stabilized zirconia (PSZ) coating was applied to the piston crown and cylinder head to establish a Low Heat Rejection (LHR) engine configuration. Experimental investigations were then conducted using diesel and CHOME biodiesel blends (B20, B40, B60, and B100) under varying engine load conditions to evaluate improvements in performance, combustion, and emission characteristics. The most important results are that the CHOME B20 blend under LHR operation produced the highest brake thermal efficiency of 34.6%, which was 8.5% higher than diesel and 12.3% higher than the conventional engine with the same fuel, while reducing brake specific fuel consumption to 0.258 kg/kWh. The LHR-B20 combination also increased peak cylinder pressure to 73 bar and maximum heat release rate to 63 J/°CA, shortened ignition delay from 11°CA to 9°CA, and reduced CO, HC, and smoke opacity emissions by 27%, 33%, and 25%, respectively, although NOx emissions increased marginally by 10%. The significance of obtained results lies in demonstrating that the integration of CHOME biodiesel with LHR engine technology can improve combustion efficiency, enhance thermal performance, and reduce harmful exhaust emissions, thereby offering a sustainable and environmentally friendly alternative fuel solution for diesel engine applications.
	Keywords: Coffee Husk Methyl Ester (CHOME), low heat rejection (LHR) engine, direct injection diesel en-gine, brake thermal efficiency, combustion characteristics, emission analysis. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.07
8	Comparative Analysis and Optimization of HVAC and Multi-Terminal HVDC Integration 400kV Super Grid (Case Study of Iraq) Authors: 1,2Sana Khalid Abdulhassan, 2Mouna Ben Smida, 3Anis Sakly 1 Al Nahrain University, Baghdad, Iraq 2University of Monastir, Monastir, Tunisia 3 ESPRIM School of Engineering, Monastir, Tunisia
	Abstract: The Iraqi 400-kV super grid faces significant operational challenges, including high transmission losses, overloaded transmission lines, voltage instability, and limited long-distance power transfer capability. The main objectives of this study are to minimize transmission losses, reduce transmission line overloading, maintain acceptable voltage profiles, and improve the overall performance of the Iraqi transmission network through the integration of Multi-Terminal High-Voltage Direct Current (MT-HVDC) technology. These objectives were achieved by modeling the Iraqi transmission network using real operational data for 2024 and analyzing it in PSS®E Version 33 under normal, 10%, and 20% loading conditions. A Genetic Algorithm (GA) was employed to identify the optimal MT-HVDC converter station locations and transmission corridors. The most important results indicate that the optimization process identified RU-MILPP, BGS4, and KESK4 as the optimal MT-HVDC terminal locations within the Iraqi 400-kV network. The proposed hybrid AC/DC configuration reduced total transmission losses by 22.5 MW and 13.75 MVar under normal operating conditions. Under a 20% load increase, the number of overloaded transmission lines decreased from 16 to 3 while maintaining all bus voltage magnitudes within the acceptable range of 0.95–1.05 p.u. Furthermore, the bipolar MT-HVDC configuration demonstrated superior loadability and long-distance power transfer capability compared with conventional HVAC transmission systems. The significance of the obtained results lies in improving transmission efficiency, enhancing voltage stability, alleviating network congestion, increasing operational flexibility, and providing a practical foundation for future renewable energy integration and regional power system interconnections.
	Keywords: HVDC, HVAC, Iraqi national grid, power losses, voltage stability, loadability, optimization, genetic algorithm, renewable energy integration. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.08
9	Integral Robustness Indicator as a Basis for Selecting a Steganographic System Under Specific Operational Conditions Authors: 1Bobok I.I., 2Kobozieva A.A. 1Odesa Polytechnic National University, Odesa, Ukraine 2Odesa National Maritime University, Odesa, Ukraine
	Abstract: Modern energy facilities are undergoing a fundamental transformation driven by the ubiquitous integration of digital technologies, the most complete embodiment of which is the Smart Grid concept. The digital transformation of energy facilities opens up new possibilities for organizing secure control data transmission channels based on steganographic principles. One of the primary requirements for a steganographic system is its robustness; however, at present, this concept in steganography is multi-criterial. For real-time systems, this complicates – or even renders impossible – the process of simultaneously accounting for or ensuring all criteria under the specific operational conditions of the steganographic system. The aim of this study is to enable the selection of a steganographic system and the a priori evaluation of its properties by developing an integral quantitative indicator of its robustness. This goal was achieved by fulfilling the following objectives: analyzing the interrelationships among existing steganographic system robustness criteria, transforming the set of robustness criteria based on these established relationships, and selecting a method for the quantitative evaluation of their priorities. The most important result of this work is the substantiation of the fundamental impossibility of creating a universal, unequivocally defined integral robustness indicator for a steganographic system, as well as substantiating the necessity of its parameterization according to operational conditions. The significance of the obtained results is illustrated using the example of organizing a secure steganographic channel for the transmission of control information within a Smart Grid infrastructure.
	Keywords: information security, Smart Grid, covert channel, steganographic system, integral metric for steganographic system robustness. DOI: https://10.52254/1857-0070.2026.3-71.09
10	Optimization of the Centralized Heating System by Redistributing Heat Loads Between the City's Combined Heat and Power Plants Authors: Tatarinova N.V., Suvorov D.M. Vyatka State University Kirov, Russian Federation
	Abstract: The aim of the study is to develop and validate methodological principles for optimizing heat load distribution between combined heat and power plants within a unified urban district heating system to enhance its energy efficiency. To achieve this goal, the following tasks were solved: the main principles and sequential stages of optimization, including preliminary analysis, strategy development, multivariate calculations, and implementation of measures, have been developed; a calculation methodology based on mathematical modeling of turbine units using their actual energy characteristics has been developed; the proposed principles have been verified on a complex of two urban CHPPs with diverse equipment, including analysis of equipment loading and compliance with temperature schedules; multivariate studies have been conducted for two fundamentally different optimization strategies at various temperature intervals; the energy effect has been assessed using the incremental efficiency criterion and integral fuel saving indicators. The most significant results con-firming the validity of the developed methodology are the following: the first strategy can be positive only under specific conditions and redistribution shares, and therefore is recommended only selectively; the second strategy yielded a considerable increase in total electricity with a substantial energy effect for all temperature intervals; the optimal redistribution degree has been determined, at which the incremental efficiency criterion reaches its minimum value. The significance of the results obtained lies in the development of a scientifically substantiated and practically applicable methodology for long-term optimization of urban district heating systems, particularly relevant amidst fuel price volatility in the modern energy market.
	Keywords: optimization, cogeneration, mathematical model, cogeneration turbine, district heating, specific heat consumption, fuel saving, energy effect. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.10
11	Investigation of the Efficiency of the Flue Gas Recirculation during Natural Gas Combustion in the Boiler Furnace Authors: Kocharyan E.V., Bedilo A.V., Arushanyan R.R., Shelest N.A. 1Kuban State Technological University, Krasnodar, Russian Federation
	Abstract: The aim of the work is to model and analyze the process of combustion of natural gas in a boiler furnace in a mixture with oxygen-enriched air while recirculating flue gases in various ways. To achieve this goal, the following objectives were addressed: the development of a mathematical model of a fire tube boiler furnace for operation during the combustion of natural gas in a mixture of air with pure oxygen and recirculated flue gases, a numerical study of the combustion process, including the temperature in the furnace and the composition of combustion products, depending on the method of recirculation. Combustion simulation was carried out in the computational fluid dynamics package OpenFOAM, using the XiFoam solver. The most important result is to obtain a formula for calculating the required degree of dryness of the recirculating flue gases, depending on the recirculation ratio and NOx emission limits. In addition, it was confirmed that the method of organizing flue gas recirculation is more efficient by feeding them in a mixture with an oxidizing agent going to combustion, compared to other methods, even under conditions of oxygen fuel combustion. The study was conducted for the recirculation ratio in the range from 0 to 50% with varying degrees of flue gas dryness. It is shown that controlling the moisture concentration in recirculated flue gases is an effective and technologically flexible tool for optimizing the natural gas combustion process in power boilers operating using oxygen combustion technology. The proposed approximation model makes it possible to estimate the required degree of dehumidification of recirculating gases at the design and commissioning stage, adapting the technology to the specific operating conditions of the power boiler and the requirements of environmental legislation.
	Keywords: oxygen combustion, fuel, emissions, recirculation, boiler, exhaust gases, furnace, flame, carbon dioxide, temperature, degree of dryness, model. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.11
12	Damping of the Cold Load Pick-up during Hourly Outage Schedules Using Gamification Methods Authors: Strunkin H.N. LLC “Pluton IC”. Zaporizhzhja, Ukraine
	Abstract: Main objectives of the study consist in the theoretical formalization and exploratory experimental evaluation of a non-capital-intensive sociotechnical demand-side management method designed for damping the parameters of a cumulative charging avalanche that occurs in low-voltage residential distribution electrical networks during scheduled blackout periods of hourly outage schedules. To demonstrate how these objectives were achieved, an analytical mathematical model of load restoration transient processes in time coordinates was synthesized based on the superposition principle and component phase decomposition, a field experimental study was performed on a typical sixty-apartment multi-family residential building with the voluntary deployment of a local demand gamification information loop, and a mathematical calculation of group simultaneity factors was carried out taking into account the integration weight coefficient of individual apartment voltage relays. The most important results are the empirical confirmation of the structural cascade splitting of household energy storage system inverter starting fronts and the controlled time shift of the mathematical expectation of the stochastic sociobehavioral demand phase, which led to a verified reduction in the cumulative peak load of the residential distribution node by 17.0% and a decrease in the recorded maximum current by 27.3 amperes under the combined mitigation scenario, successfully minimizing the risk of accelerated thermal degradation of input fuse links. The significance of obtained results lies in the creation of an efficient, scalable, and non-invasive engineering tool for dis-tribution system operators, the potential integration of which into mobile customer services and automated supervisory control complexes ensures local feeder stability and transformer load profile leveling without expensive modernization of physical equipment and grid cable infrastructure.
	Keywords: hourly outage schedules, energy storage systems, Cold Load Pickup, “charging avalanche”, cascade decomposition, gamification. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.12
13	Peculiarities of Thermal Regimes and Emission Characteristics of Hot Wa-ter Boilers Operating on Heterogeneous Fuel Pellets Authors: Rachynskyi A.Yu., Chernousenko O.Yu., Baraniuk O.V., Onysko A.I., Tkachenko S.S. National Technical University of Ukraine «Igor Sikorsky Kyiv Polytechnic Institute» Kyiv, Ukraine
	Abstract: The aim of this study is to provide a scientific substantiation of the mechanisms of thermochemical conversion of wood pellets and low-grade agrobiomass in the furnace chambers of municipal boilers using multiparameter CFD modeling methods. To achieve this aim, the following tasks were solved: a numerical approach combining the Discrete Phase Method (DPM) and the Laminar Flamelet mod-el was implemented to account for the staged release of volatile matter; the influence of intensive aerodynamic turbulence on particle trajectories was investigated; and the processes of staged air sup-ply were modeled. The scientific novelty of the results lies in establishing the regularities of local reducing zone formation in the flame root during the combustion of high-nitrogen fuels (bottom sludge and poultry manure), for which it was theoretically demonstrated that the vortex effect of turbulence ensures a burnout completeness of the dispersed phase at the level of 98.0–99.2%. The most important results are the proof that staged air supply forms stable reducing zones near the burner, limiting the local NOx concentration to 36 ppm, as well as the mathematical determination of the safe slag-free interval (1000–1150 °C) for the thermal conversion of straw and rapeseed. The practical value of the work consists in the verification of the numerical model based on detailed chemical kinetics approaches, which provides increased accuracy in predicting reducing zones com-pared with standard Eddy Dissipation models. The significance of the obtained results lies in con-firming the possibility of efficient and environmentally safe utilization of low-quality agrobiomass and industrial sludge in serial boiler units. This determines the optimal operating conditions of the equipment and creates a fundamental scientific basis for diversification of the fuel base and exten-sion of the residual service life of thermal power installations under deep load maneuvering.
	Keywords: fuel pellets, bioenergy, flame aerodynamics, heat and mass transfer, waste utilization, model verifi-cation, environmental safety, surface slagging, energy efficiency. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.13
14	Evaluation of Operating Modes for Central Heating Stations Using Neural Networks and Ensemble Machine Learning Methods Authors: Dvortsevoy A.I., Borush O.V., Yakovina I.N., Boyko E.E., Vindemut D.Y., Yankov P.A. Novosibirsk State Technical University Novosibirsk, Russian Federation
	Abstract: The main objective of this study is to develop a method for assessing the operating modes of central heating stations (CHS) based on forecasting the thermal power and analyzing the deviations of actual parameters from the reference values. To achieve this goal, the following tasks were solved: primary processing and cleaning of data on technological parameters of CHS and meteorological observations for the period 2022-2024 were performed; regression models (Random Forest, XGBoost, CatBoost and a fully connected neural network) were trained and compared for forecasting the thermal power; an algorithm for assessing and forecasting the operating modes of CHS was developed. The most important results are: the fully connected neural network model, which achieved the determination coefficient R² = 0.9997 and the mean absolute error MAE = 0.0099 Gcal/h; An algorithm developed based on the resulting model allows for differentiating deviations in central heating system operating modes by their source—from the heat supplier, from the consumer, and hidden abnormal conditions not detected by direct comparison with the temperature chart. The analysis revealed that a sharp increase in outside air temperature leads to an increase in the direct network water pressure due to the specifics of quantitative weather-dependent regulation, which can cause accelerated wear of heating networks. The significance of the obtained results lies in the fact that the proposed algorithm ensures the early detection of emergency and preemergency conditions, which helps reduce accident rates, reduce repair and maintenance costs, and improve the energy efficiency of district heating systems. The method can be adapted to various regions with centralized heating.
	Keywords: district heating substation, machine learning, neural network forecasting, regression analysis, anomaly detection, thermal power, temperature schedule, weather-dependent control, pre-emergency mode, heat supply monitoring. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.14
15	Computer Vision Capabilities for Detecting and Classifying Solar Panel Faults Authors: 1Mohammed N.S., 1Uonis M.M., 2Alsaif O.I. 1Mosul University Mosul, Iraq 2Northern Technical University Polytechnique College-Mosul Mosul, Iraq
	Abstract: The main objectives of this research are to fill a gap in PV inspection systems, which consist of convolutional detectors with a limited amount of global context or transformer classifiers with weak localization. Defects micro-cracks, broken electrodes, and degradation regions decrease the PV energy efficiency considerably, and this encourages the use of an automated computer vision algorithm for the classification of these defects extracted from electroluminescence (EL) images. These objectives were achieved by solving the following task: a novel dual-stage architecture is used to solve these tasks, which are performed sequentially, YOLOv8m followed by the Swin Transformer architecture. In comparison with previous CNN-Transformer PV models that combine features by concatenation or averaging with predefined weights, the proposed model implements a learnable cross-attention fusion probability merging ensemble, which dynamically weights the localization and global attention evidence for each defect. YOLOv8m localizes the exact positions of defects, then the Swin Transformer analyzes the region by window-swiping self-attention to return a long-range degradation pattern that CNNs can't pass. The model was trained using EL images in 12 classes, such as, intact cells, cracks, point defects, finger damage, and EL images were augmented to overcome the class imbalance problem and cross-conditioning generalization. The most important results are a macro-averaged F1-score of 94.05% and 95.40% accuracy, with per-class F1 ranging from 98.98% (Mono-crystalline) to 89.04% (Micro-crack). The significance of the obtained results lies in the use of integrative localization-driven YOLOv8m convolutional features with transformer global attention and learnable cross-attention fusion instead of static cross-attention fusion, providing a new and more robust implementation to perform automated PV quality assessment than a single branch.
	Keywords: cross-attention fusion (CAF), computer vision, deep learning, electroluminescence imaging (EL), ensemble deep learning, solar cell defect classification, swin-transformer (ST), YOLOv8m. DOI: https://doi.org/10.52254/1857-0070.2026.3-71.15