Dmytro Danylchenko
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Dmytro Danylchenko has a total of 100 co-authors. Here are the top co-authors by number of shared publications:
- Sergii Shevchenko
National Technical University “Kharkiv Polytechnic Institute
19 shared publications - Stanislav Dryvetskyi
National Technical University “Kharkiv Polytechnic Institute
12 shared publications - Stanislav Fedorchuk
National Technical University, Kharkiv Polytechnic Institute,dept. Electric Power Stations,Kharkiv,Ukraine
10 shared publications - Andrii Ivakhnov
National Technical University, Kharkiv Polytechnic Institute,dept. Electric Power Stations,Kharkiv,Ukraine
9 shared publications - Andrii Potryvai
Національний технічний університет «Харківський політехнічний інститут», Україна image/svg+xml
8 shared publications - Oleksandr Kulapin
National Technical University, Kharkiv Polytechnic Institute,dept. Electric Power Stations,Kharkiv,Ukraine
7 shared publications
Latest Publications
- Open Access
- 30 September 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 77-83
Application features of flat busbars and tubular conductors in underground utility tunnels
Purpose. To assess the feasibility of using a flat busbar and a hollow cylindrical conductor with an unchanged aluminium cross-sectional area of 240 mm² in a compact high-voltage line in an underground collector. Selecting the phase-conductor shape requires coordination between cooling conditions and electrical parameters. Methodology. A vertical 80×3 mm busbar is compared with tubes having different outer diameters within the range of 30–40 mm. The study uses geometric calculations, a dimensionless-correlation model of natural convection, and an analytical model of a transposed three-phase line with a 300 mm spacing between adjacent phase axes. The air and conductor temperatures are taken as 15 and 70 °C, respectively. For the flat busbar, the electrostatic radius in the thin-strip approximation is distinguished from the geometric mean distance for a uniform current density. Findings. The tube perimeter increases from 94.25 to 125.66 mm but remains below the 166 mm perimeter of the flat busbar. The wall thickness decreases from 2.810 to 2.011 mm. The calculated tube capacitance increases from 17.24 to 18.93 nF/km, while surge impedance loading increases from 61.95 to 68.27 MW. The flat busbar yields 67.80 MW. However, the small difference relative to the 40 mm tube does not allow the alternatives to be ranked confidently without refining the field calculation. Originality. The methodology is developed through a joint comparison of the heat-transfer surface and electrical characteristics of cross-sections containing the same amount of material, with an explicit distinction between the assumptions used. The electrostatic radius and geometric mean distance are treated separately. Practical value. The results are suitable for preliminary selection of conductor geometry. A final permissible-current assessment requires the AC resistance, surface condition, and a thermal model of the collector.
- 30 September 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 77-83
Open AccessApplication features of flat busbars and tubular conductors in underground utility tunnels
Purpose. To assess the feasibility of using a flat busbar and a hollow cylindrical conductor with an unchanged aluminium cross-sectional area of 240 mm² in a compact high-voltage line in an underground collector. Selecting the phase-conductor shape requires coordination between cooling conditions and electrical parameters. Methodology. A vertical 80×3 mm busbar is compared with tubes having different outer diameters within the range of 30–40 mm. The study uses geometric calculations, a dimensionless-correlation model of natural convection, and an analytical model of a transposed three-phase line with a 300 mm spacing between adjacent phase axes. The air and conductor temperatures are taken as 15 and 70 °C, respectively. For the flat busbar, the electrostatic radius in the thin-strip approximation is distinguished from the geometric mean distance for a uniform current density. Findings. The tube perimeter increases from 94.25 to 125.66 mm but remains below the 166 mm perimeter of the flat busbar. The wall thickness decreases from 2.810 to 2.011 mm. The calculated tube capacitance increases from 17.24 to 18.93 nF/km, while surge impedance loading increases from 61.95 to 68.27 MW. The flat busbar yields 67.80 MW. However, the small difference relative to the 40 mm tube does not allow the alternatives to be ranked confidently without refining the field calculation. Originality. The methodology is developed through a joint comparison of the heat-transfer surface and electrical characteristics of cross-sections containing the same amount of material, with an explicit distinction between the assumptions used. The electrostatic radius and geometric mean distance are treated separately. Practical value. The results are suitable for preliminary selection of conductor geometry. A final permissible-current assessment requires the AC resistance, surface condition, and a thermal model of the collector.
- Open Access
- 30 September 2026
- Published by Politechnika Lubelska in Applied Computer Science
- Vol. 22 (3) , 82-96
Digitalisation of relay protection and implementation of main protections in digital form
This paper presents a Java-based software demonstration of overcurrent, transformer differential, and distance protection. Its contribution is an implementation-oriented treatment of three complementary protection functions, with explicit signal-processing, decision, and timing logic. The transformer differential component builds on the authors’ earlier study; the present scope also includes four-stage overcurrent logic and a comparison of polygonal and circular distance characteristics. The methodology combines a scalar overcurrent example with offline processing of fault records for the other functions. Measurements of the Java implementation yielded approximately 30–70 ms operating time, 5–20 µs processing time per sample, and a JVM memory footprint of approximately 30–80 MB. At the adopted 4 kHz sampling rate, the reported per-sample processing time corresponds to 2–8% of the 250 µs sampling interval. These measurements support computational feasibility under the tested conditions, while qualitative timing variability remains a limitation. The results do not establish a worst-case execution bound, statistical reliability, or compliance with protection-equipment standards.
- 30 September 2026
- Published by Politechnika Lubelska in Applied Computer Science
- Vol. 22 (3) , 82-96
Open AccessDigitalisation of relay protection and implementation of main protections in digital form
This paper presents a Java-based software demonstration of overcurrent, transformer differential, and distance protection. Its contribution is an implementation-oriented treatment of three complementary protection functions, with explicit signal-processing, decision, and timing logic. The transformer differential component builds on the authors’ earlier study; the present scope also includes four-stage overcurrent logic and a comparison of polygonal and circular distance characteristics. The methodology combines a scalar overcurrent example with offline processing of fault records for the other functions. Measurements of the Java implementation yielded approximately 30–70 ms operating time, 5–20 µs processing time per sample, and a JVM memory footprint of approximately 30–80 MB. At the adopted 4 kHz sampling rate, the reported per-sample processing time corresponds to 2–8% of the 250 µs sampling interval. These measurements support computational feasibility under the tested conditions, while qualitative timing variability remains a limitation. The results do not establish a worst-case execution bound, statistical reliability, or compliance with protection-equipment standards.
- Open Access
- 1 September 2026
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 3329 (1) , 012005
Modeling and optimization of rural distribution networks with solar power plants and energy storage systems
This paper addresses the modeling and optimization of operating modes of rural distribution networks integrated with solar power plants (SPPs) and energy storage systems (ESS). The relevance of the study is determined by the increasing penetration of renewable energy sources, the need to enhance energy security, and the requirement to ensure reliable electricity supply for rural consumers under unstable grid conditions. The aim of the research is to develop a mathematical modeling approach for a local distribution network with integrated SPP and ESS and to determine efficient charge–discharge control strategies for reducing peak loads, minimizing power losses, and increasing the level of energy autonomy. The study applies simulation-based modeling methods, generation forecasting using meteorological data, and mathematical optimization techniques to form optimal ESS operating schedules. The results demonstrate that the integration of ESS into rural distribution networks significantly reduces load fluctuations, increases self-consumption of locally generated electricity, decreases grid power exchange, and improves overall system stability. Technical and economic performance indicators of ESS implementation for rural facilities, including pumping stations and agricultural loads, are evaluated.
- 1 September 2026
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 3329 (1) , 012005
Open AccessModeling and optimization of rural distribution networks with solar power plants and energy storage systems
This paper addresses the modeling and optimization of operating modes of rural distribution networks integrated with solar power plants (SPPs) and energy storage systems (ESS). The relevance of the study is determined by the increasing penetration of renewable energy sources, the need to enhance energy security, and the requirement to ensure reliable electricity supply for rural consumers under unstable grid conditions. The aim of the research is to develop a mathematical modeling approach for a local distribution network with integrated SPP and ESS and to determine efficient charge–discharge control strategies for reducing peak loads, minimizing power losses, and increasing the level of energy autonomy. The study applies simulation-based modeling methods, generation forecasting using meteorological data, and mathematical optimization techniques to form optimal ESS operating schedules. The results demonstrate that the integration of ESS into rural distribution networks significantly reduces load fluctuations, increases self-consumption of locally generated electricity, decreases grid power exchange, and improves overall system stability. Technical and economic performance indicators of ESS implementation for rural facilities, including pumping stations and agricultural loads, are evaluated.
- Open Access
- 29 May 2026
- Published by National Technical University Kharkiv Polytechnic Institute in Bulletin of the National Technical University "KhPI". Series: Energy: Reliability and Energy Efficiency
- No. 1(12),p. 48-52
Simulation of electromagnetic processes in the power supply of induction motors from underground substations
The article considers the influence of cable line parameters on the power supply of induction motors from underground electrical substations. Ensuring reliable starting of induction motors from underground transformer substations is a critical task. The significant remoteness of consumers necessitates the use of long cable lines, which leads to deep voltage drops during direct start. The presence of nonlinear loads in the network that generate higher current and voltage harmonics further worsens energy performance by increasing total electrical and magnetic power losses in motors and transformers, accelerated thermal aging and insulation degradation in conditions of high humidity and aggressive environments, reduced dynamic stability of electric drives, and risks of process disruption. n this study, a comprehensive simulation of the power supply of induction motors from local and central underground electrical substations without and with the presence of higher voltage harmonics was carried out. An analysis of the voltage drop depending on the length and cross-section of the cable cores for the 1140 V network during direct start of an induction motor for capacities of 100–132 kW was performed, and an assessment of the influence of the harmonic composition of the supply voltage on the maximum (maximum) length of the network at which the conditions for normal direct start at the nominal torque of the induction motor are ensured was also performed. It has been established that in the presence of higher harmonics, the starting current increases, and therefore the voltage drop increases by an additional 4–6 %, which requires a forced reduction in the calculated network length by 15–20 %. It has been proven that to overcome these limitations, when powering an induction motor at distances of more than 1 km from an underground substation, it is necessary to abandon direct starting in favor of using soft starting devices (thyristor voltage regulators) or local vector frequency control systems.
- 29 May 2026
- Published by National Technical University Kharkiv Polytechnic Institute in Bulletin of the National Technical University "KhPI". Series: Energy: Reliability and Energy Efficiency
- No. 1(12),p. 48-52
Open AccessSimulation of electromagnetic processes in the power supply of induction motors from underground substations
The article considers the influence of cable line parameters on the power supply of induction motors from underground electrical substations. Ensuring reliable starting of induction motors from underground transformer substations is a critical task. The significant remoteness of consumers necessitates the use of long cable lines, which leads to deep voltage drops during direct start. The presence of nonlinear loads in the network that generate higher current and voltage harmonics further worsens energy performance by increasing total electrical and magnetic power losses in motors and transformers, accelerated thermal aging and insulation degradation in conditions of high humidity and aggressive environments, reduced dynamic stability of electric drives, and risks of process disruption. n this study, a comprehensive simulation of the power supply of induction motors from local and central underground electrical substations without and with the presence of higher voltage harmonics was carried out. An analysis of the voltage drop depending on the length and cross-section of the cable cores for the 1140 V network during direct start of an induction motor for capacities of 100–132 kW was performed, and an assessment of the influence of the harmonic composition of the supply voltage on the maximum (maximum) length of the network at which the conditions for normal direct start at the nominal torque of the induction motor are ensured was also performed. It has been established that in the presence of higher harmonics, the starting current increases, and therefore the voltage drop increases by an additional 4–6 %, which requires a forced reduction in the calculated network length by 15–20 %. It has been proven that to overcome these limitations, when powering an induction motor at distances of more than 1 km from an underground substation, it is necessary to abandon direct starting in favor of using soft starting devices (thyristor voltage regulators) or local vector frequency control systems.
- Open Access
- 29 May 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 2,p. 67-73
Assessment of the effect of structural and geophysical parameters on the magnetic field distribution around power lines located in underground collectors
Purpose. To quantitatively estimate the contribution of the surrounding-medium parameters to the magnetic flux density of a three-phase 110 kV air-insulated transmission line installed inside an underground tunnel, and to clarify whether such parameters must be included in routine engineering magnetic-field assessments. Methodology. Numerical quasi-stationary simulation of the magnetic field at industrial frequency was carried out for a symmetric three-phase system of AC-240/32 conductors loaded with 500 A. The computational domain was decomposed into five characteristic horizontal layers extending from the open air above the ground surface to the very plane of the phase wires. Two parametric series were performed: in the first the soil resistivity was scanned from 50 to 1000 Ω·m at a fixed wall thickness; in the second the concrete-wall thickness was varied from 250 to 500 mm at a fixed soil resistivity of 70 Ω·m. Findings. Within the studied envelope of parameters neither the resistivity of the soil nor the thickness of the concrete wall measurably alters the magnetic flux density either at the control point or inside any of the analyzed layers. The configuration of the field is governed almost exclusively by the magnitude of the phase current and by the geometry of the conductor cluster. At the reference point the peak flux density is approximately 1.5 μT and is reached at phase angles of about 115° and 295°; the value rises to roughly 2.27 μT at the ground surface, to about 8 μT on the upper outer face of the tunnel and exceeds 10 μT on its inner face, while pronounced local maxima are observed in the immediate vicinity of the individual phase wires. Originality. A systematic quantitative demonstration is provided that for a 110 kV underground concrete collector in typical urban operating conditions the soil and the reinforced-concrete envelope contribute negligibly to magnetic-field attenuation at 50 Hz. The widespread informal design assumption about the natural shielding role of these media is therefore not supported within the considered ranges of parameters. Practical value. The reported results allow designers to omit detailed soil and concrete characterization from preliminary magnetic-field calculations above underground collectors and to instead direct mitigation efforts toward optimization of phase-conductor geometry or toward dedicated active and passive shielding solutions, the only practically effective routes to a noticeable reduction of magnetic flux density in compact underground transmission infrastructure.
- 29 May 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 2,p. 67-73
Open AccessAssessment of the effect of structural and geophysical parameters on the magnetic field distribution around power lines located in underground collectors
Purpose. To quantitatively estimate the contribution of the surrounding-medium parameters to the magnetic flux density of a three-phase 110 kV air-insulated transmission line installed inside an underground tunnel, and to clarify whether such parameters must be included in routine engineering magnetic-field assessments. Methodology. Numerical quasi-stationary simulation of the magnetic field at industrial frequency was carried out for a symmetric three-phase system of AC-240/32 conductors loaded with 500 A. The computational domain was decomposed into five characteristic horizontal layers extending from the open air above the ground surface to the very plane of the phase wires. Two parametric series were performed: in the first the soil resistivity was scanned from 50 to 1000 Ω·m at a fixed wall thickness; in the second the concrete-wall thickness was varied from 250 to 500 mm at a fixed soil resistivity of 70 Ω·m. Findings. Within the studied envelope of parameters neither the resistivity of the soil nor the thickness of the concrete wall measurably alters the magnetic flux density either at the control point or inside any of the analyzed layers. The configuration of the field is governed almost exclusively by the magnitude of the phase current and by the geometry of the conductor cluster. At the reference point the peak flux density is approximately 1.5 μT and is reached at phase angles of about 115° and 295°; the value rises to roughly 2.27 μT at the ground surface, to about 8 μT on the upper outer face of the tunnel and exceeds 10 μT on its inner face, while pronounced local maxima are observed in the immediate vicinity of the individual phase wires. Originality. A systematic quantitative demonstration is provided that for a 110 kV underground concrete collector in typical urban operating conditions the soil and the reinforced-concrete envelope contribute negligibly to magnetic-field attenuation at 50 Hz. The widespread informal design assumption about the natural shielding role of these media is therefore not supported within the considered ranges of parameters. Practical value. The reported results allow designers to omit detailed soil and concrete characterization from preliminary magnetic-field calculations above underground collectors and to instead direct mitigation efforts toward optimization of phase-conductor geometry or toward dedicated active and passive shielding solutions, the only practically effective routes to a noticeable reduction of magnetic flux density in compact underground transmission infrastructure.
- 21 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Prosumer as the Smallest Energy Hub and the Prerequisites for its Operation at a Decentralized Power Grid
Modern global energy trends are characterized by a paradigm shift from traditional centralized management toward decentralized architectures. This transition involves incorporating distributed generation and structuring local networks into functional energy hubs, which serves to mitigate the operational challenges imposed by the volatile nature of intermittent renewable energy sources. This study explores the structural layout of energy hubs, focusing specifically on prosumers as the fundamental, granular units within decentralized power systems.
- 21 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Prosumer as the Smallest Energy Hub and the Prerequisites for its Operation at a Decentralized Power Grid
Modern global energy trends are characterized by a paradigm shift from traditional centralized management toward decentralized architectures. This transition involves incorporating distributed generation and structuring local networks into functional energy hubs, which serves to mitigate the operational challenges imposed by the volatile nature of intermittent renewable energy sources. This study explores the structural layout of energy hubs, focusing specifically on prosumers as the fundamental, granular units within decentralized power systems.
- Open Access
- 30 March 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 1,p. 7-15
Improving approaches to the linearization of magnetic properties of electromechanical converter cores for numerical calculations
Purpose. Improving the mathematical model for the linearization of magnetic properties of stator and rotor electrical steel based on the piecewise linear approximation of the magnetization curve, which will significantly reduce the number of nonlinear equations and substantially decrease the computational time for 2D and 3D circuit-field models of an asynchronous electromechanical converter while maintaining the accuracy of numerical calculations. Methodology. Analytical methods for electromagnetic field calculation, the finite element method, analytical methods for electrical and magnetic circuit calculation, and piecewise linear approximation methods are utilized. Findings. Mathematical models for linearizing the magnetization curve of stator and rotor electrical steel has been improved for both spatial 3D and plane-parallel 2D geometric models of the active part of AC electromechanical converters. An approach has been implemented to partition the general 2D geometric model of the stator and rotor core into n-elementary planes, applying a linear magnetization curve function at µr = const. Based on the 2D field modeling results of an electromechanical converters prototype in idle mode, compared to modeling results considering the full magnetization curve across the entire computational domain, the armature current residual does not exceed 2.654%. This allowed for a 4.89-fold reduction in the numerical calculation time for the 2D transient problem formulation using Comsol Multiphysics software. Originality. The method for linearizing the magnetic properties of stator and rotor electrical steel in electromechanical converters has been improved, allowing for the application of a constant relative permeability value (µr =const.) to specific areas of the stator and rotor core computational domain while maintaining a sufficient level of numerical calculation accuracy. Practical value. The proposed methodology for the linearization of the magnetization curve of stator and rotor electrical steel for spatial 3D and plane-parallel 2D geometric models of the active part of AC electromechanical converters can be applied to various types of electrical machines.
- 30 March 2026
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 1,p. 7-15
Open AccessImproving approaches to the linearization of magnetic properties of electromechanical converter cores for numerical calculations
Purpose. Improving the mathematical model for the linearization of magnetic properties of stator and rotor electrical steel based on the piecewise linear approximation of the magnetization curve, which will significantly reduce the number of nonlinear equations and substantially decrease the computational time for 2D and 3D circuit-field models of an asynchronous electromechanical converter while maintaining the accuracy of numerical calculations. Methodology. Analytical methods for electromagnetic field calculation, the finite element method, analytical methods for electrical and magnetic circuit calculation, and piecewise linear approximation methods are utilized. Findings. Mathematical models for linearizing the magnetization curve of stator and rotor electrical steel has been improved for both spatial 3D and plane-parallel 2D geometric models of the active part of AC electromechanical converters. An approach has been implemented to partition the general 2D geometric model of the stator and rotor core into n-elementary planes, applying a linear magnetization curve function at µr = const. Based on the 2D field modeling results of an electromechanical converters prototype in idle mode, compared to modeling results considering the full magnetization curve across the entire computational domain, the armature current residual does not exceed 2.654%. This allowed for a 4.89-fold reduction in the numerical calculation time for the 2D transient problem formulation using Comsol Multiphysics software. Originality. The method for linearizing the magnetic properties of stator and rotor electrical steel in electromechanical converters has been improved, allowing for the application of a constant relative permeability value (µr =const.) to specific areas of the stator and rotor core computational domain while maintaining a sufficient level of numerical calculation accuracy. Practical value. The proposed methodology for the linearization of the magnetization curve of stator and rotor electrical steel for spatial 3D and plane-parallel 2D geometric models of the active part of AC electromechanical converters can be applied to various types of electrical machines.
- Open Access
- 30 December 2025
- Published by National Technical University Kharkiv Polytechnic Institute in Bulletin of the National Technical University "KhPI". Series: Energy: Reliability and Energy Efficiency
- No. 2(11),p. 18-24
Application of artificial neural networks in relay protection systems of power supply networks for fault condition recognition
The paper investigates the possibility of applying artificial neural networks in relay protection systems of power supply networks in order to improve the reliability, speed, and adaptability of recognizing emergency and pre-emergency operating conditions. It is shown that conventional relay protection devices based on fixed settings and logical algorithms have limited capability to adapt to changes in power system parameters, network topology, and load conditions, which may lead to false or unnecessary operations. To address this problem, a neural network-based approach is proposed, relying on the analysis of instantaneous values of phase currents and voltages. A simulation model of the power supply system was developed in the MATLAB/Simulink environment using standard electrical component libraries. During the simulation process, normal and fault operating modes were reproduced, including conditions caused by single-phase short circuits, which enabled the formation of representative training and testing datasets. The paper presents the structure of the artificial neural network, describes the training process using the backpropagation algorithm, and analyzes the convergence of the training procedure based on the mean squared error criterion. Testing of the trained neural network model was performed, and the retraining process under changing power system parameters was also investigated. The obtained results confirm the ability of the neural network to effectively recognize operating modes of the power supply system and to maintain correct performance after retraining. It is concluded that the application of artificial neural networks as an intelligent element of relay protection systems is a promising approach for increasing their adaptability, reducing the probability of false operations, and ensuring stable operation of power systems under varying operating conditions.
- 30 December 2025
- Published by National Technical University Kharkiv Polytechnic Institute in Bulletin of the National Technical University "KhPI". Series: Energy: Reliability and Energy Efficiency
- No. 2(11),p. 18-24
Open AccessApplication of artificial neural networks in relay protection systems of power supply networks for fault condition recognition
The paper investigates the possibility of applying artificial neural networks in relay protection systems of power supply networks in order to improve the reliability, speed, and adaptability of recognizing emergency and pre-emergency operating conditions. It is shown that conventional relay protection devices based on fixed settings and logical algorithms have limited capability to adapt to changes in power system parameters, network topology, and load conditions, which may lead to false or unnecessary operations. To address this problem, a neural network-based approach is proposed, relying on the analysis of instantaneous values of phase currents and voltages. A simulation model of the power supply system was developed in the MATLAB/Simulink environment using standard electrical component libraries. During the simulation process, normal and fault operating modes were reproduced, including conditions caused by single-phase short circuits, which enabled the formation of representative training and testing datasets. The paper presents the structure of the artificial neural network, describes the training process using the backpropagation algorithm, and analyzes the convergence of the training procedure based on the mean squared error criterion. Testing of the trained neural network model was performed, and the retraining process under changing power system parameters was also investigated. The obtained results confirm the ability of the neural network to effectively recognize operating modes of the power supply system and to maintain correct performance after retraining. It is concluded that the application of artificial neural networks as an intelligent element of relay protection systems is a promising approach for increasing their adaptability, reducing the probability of false operations, and ensuring stable operation of power systems under varying operating conditions.
- Open Access
- 22 October 2025
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 52-60
Way to implementation in relay protection of the algorithm for searching for pre-fault operating conditions by using a neural network with backpropagation spot training
Purpose. The purpose of this paper is to develop and validate a novel method for searching critical and pre-fault operating conditions in power systems for relay protection setting calculation. The proposed approach is based on the integration of graph neural networks with deep reinforcement learning (GNN-DRL) and aims to significantly reduce computational time while maintaining high accuracy and ensuring protection selectivity. Findings. It is demonstrated that the proposed GNN-DRL method achieves a 10–1000 times reduction in computational overhead compared to traditional brute-force approaches. The prediction accuracy exceeds 90%, while e-accuracy surpasses 98%. The method proves to be highly effective both in determining maximum fault currents for instantaneous overcurrent protection and in large-scale scenarios on extended test systems. Originality. For the first time, the integration of graph neural networks with the Dueling Double Deep Q Network algorithm is proposed to solve the problem of searching for critical and pre-fault operating conditions in relay protection. A unique two-stage training framework, GLFE, is developed, combining elements of supervised and unsupervised learning. This enables high prediction accuracy while requiring significantly fewer training samples. Practical value. The results of this study can be applied in the design of intelligent relay protection and automation systems for electric power networks. The proposed method is capable of adapting to fast-changing conditions in power systems with high penetration of renewable energy. Furthermore, the architecture can be extended to other optimization and control problems in complex energy environments.
- 22 October 2025
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 52-60
Open AccessWay to implementation in relay protection of the algorithm for searching for pre-fault operating conditions by using a neural network with backpropagation spot training
Purpose. The purpose of this paper is to develop and validate a novel method for searching critical and pre-fault operating conditions in power systems for relay protection setting calculation. The proposed approach is based on the integration of graph neural networks with deep reinforcement learning (GNN-DRL) and aims to significantly reduce computational time while maintaining high accuracy and ensuring protection selectivity. Findings. It is demonstrated that the proposed GNN-DRL method achieves a 10–1000 times reduction in computational overhead compared to traditional brute-force approaches. The prediction accuracy exceeds 90%, while e-accuracy surpasses 98%. The method proves to be highly effective both in determining maximum fault currents for instantaneous overcurrent protection and in large-scale scenarios on extended test systems. Originality. For the first time, the integration of graph neural networks with the Dueling Double Deep Q Network algorithm is proposed to solve the problem of searching for critical and pre-fault operating conditions in relay protection. A unique two-stage training framework, GLFE, is developed, combining elements of supervised and unsupervised learning. This enables high prediction accuracy while requiring significantly fewer training samples. Practical value. The results of this study can be applied in the design of intelligent relay protection and automation systems for electric power networks. The proposed method is capable of adapting to fast-changing conditions in power systems with high penetration of renewable energy. Furthermore, the architecture can be extended to other optimization and control problems in complex energy environments.
- Open Access
- 22 October 2025
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 38-43
Specific features of modeling distributed generation within electrical network structures using different software packages
Purpose. The purpose of this study is to investigate the specific features of modeling distributed generation in the structure of electrical networks using different software packages. The work aims to identify the software that provides the necessary functionality, accuracy of calculations, ease of use, and license availability, while also considering the specific characteristics of Ukrainian electrical networks. Methodology. Analysis of literary sources — applied to study the current state of the problem of distributed generation integration into electrical networks and to summarize both international and Ukrainian experience. Comparative analysis — carried out to evaluate the functional capabilities of PowerFactory, Dakar Eleks, Matlab/Simulink, and Ansys according to the following criteria: modeling accuracy, computational performance, interface usability, availability of model libraries of distributed generation, the ability to model active and reactive power characteristics, voltage control characteristics, synchronization modes, and transient processes. Findings. The key technical aspects of connecting distributed generation were identified, including the modeling of active and reactive power characteristics, voltage control characteristics, synchronization modes, transient processes, and automatic reconnection in accordance with the requirements of the Distribution System Code and the Transmission System Code. A comparative analysis of the functional capabilities of PowerFactory, Dakar Eleks, Matlab/Simulink, and Ansys was performed. International experience with the use of these software tools and the features of their adaptation to Ukrainian regulatory requirements were also analyzed. Originality. A new approach is proposed for evaluating software packages for modeling the connection of distributed generation, which takes into account the specific features of Ukrainian electrical networks and the requirements of the Distribution System Code and the Transmission System Code. Practical value. The results of this study can be applied in the process of selecting software for modeling the integration of distributed generation in Ukrainian electrical networks. The proposed approach makes it possible to reasonably determine optimal connection points, assess the influence of distributed generation on the operating conditions of the power system, and improve the efficiency of network development planning.
- 22 October 2025
- Published by National University "Zaporizhzhia Polytechnic" in Electrical Engineering and Power Engineering
- No. 3,p. 38-43
Open AccessSpecific features of modeling distributed generation within electrical network structures using different software packages
Purpose. The purpose of this study is to investigate the specific features of modeling distributed generation in the structure of electrical networks using different software packages. The work aims to identify the software that provides the necessary functionality, accuracy of calculations, ease of use, and license availability, while also considering the specific characteristics of Ukrainian electrical networks. Methodology. Analysis of literary sources — applied to study the current state of the problem of distributed generation integration into electrical networks and to summarize both international and Ukrainian experience. Comparative analysis — carried out to evaluate the functional capabilities of PowerFactory, Dakar Eleks, Matlab/Simulink, and Ansys according to the following criteria: modeling accuracy, computational performance, interface usability, availability of model libraries of distributed generation, the ability to model active and reactive power characteristics, voltage control characteristics, synchronization modes, and transient processes. Findings. The key technical aspects of connecting distributed generation were identified, including the modeling of active and reactive power characteristics, voltage control characteristics, synchronization modes, transient processes, and automatic reconnection in accordance with the requirements of the Distribution System Code and the Transmission System Code. A comparative analysis of the functional capabilities of PowerFactory, Dakar Eleks, Matlab/Simulink, and Ansys was performed. International experience with the use of these software tools and the features of their adaptation to Ukrainian regulatory requirements were also analyzed. Originality. A new approach is proposed for evaluating software packages for modeling the connection of distributed generation, which takes into account the specific features of Ukrainian electrical networks and the requirements of the Distribution System Code and the Transmission System Code. Practical value. The results of this study can be applied in the process of selecting software for modeling the integration of distributed generation in Ukrainian electrical networks. The proposed approach makes it possible to reasonably determine optimal connection points, assess the influence of distributed generation on the operating conditions of the power system, and improve the efficiency of network development planning.