R. B. Bagmut
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
R. B. Bagmut has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- G. A. Pliekhova
Kharkiv National Automobile and Highway University
3 shared publications - Maryna Kostikova
ХНАДУ
3 shared publications - Danylo Pliekhov
Kharkiv National Automobile and Highway University
2 shared publications - Igor Gamayun
National Technical University "Kharkiv Polytechnic Institute
2 shared publications - Andrii Shyshatskyi
Kharkiv National Automobile and Highway University
1 shared publication - Andrii Veretnov
Central Scientifically-Research Institute of Armaments and Military Equipment of the Armed Forces of Ukraine
1 shared publication
Latest Publications
- Open Access
- 22 October 2026
- Published by Private Company Technology Center in Eastern-European Journal of Enterprise Technologies
- Vol. 5 (4 (143)) , 58-67
Development of a comprehensive method of learning artificial neural networks for organizational and technical systems
The object of the study is the learning process of artificial neural networks (ANNs). The problem addressed in the study is to increase the reliability of ANN learning while ensuring a given level of learning efficiency. The originality of the proposed complex method is: – verification of the architecture of organizational and technical systems (OTS) and the dynamics of its change, due to the use of an improved cheetah pack algorithm. This allows leveling the influence of the human factor and increasing the reliability of ANN learning; – taking into account the possible values of the state of the environment and the actions of agents, which allows taking into account in full the conditions and factors in which ANN learning is carried out. Due to this, the reliability of the ANN learning process increases under the influence of destabilizing factors; – formation of a generalized architecture of multi agent learning due to the use of the ensemble mutation procedure. Due to the use of this procedure, it becomes possible to achieve an increase in the reliability of ANN learning; – selected the type and parameters of the membership function due to the use of the appropriate procedure. Due to the use of this procedure, it becomes possible to achieve an increase in the accuracy of ANN learning; – reduction of sensitivity to a priori hyperparameters of models and tendency to excessive fragmentation of data in OTS, due to the use of improved procedures, which increases the reliability of ANN learning; – for ANN learning, a sample with an uneven distribution of the states of OTS functioning is used, which achieves the flexibility of ANN learning under the influence of destabilizing factors. The proposed complex method of ANN learning minimizes the geometric error of the division by an average of 25%, thereby increasing the reliability of ANN learning while maintaining a given level of efficiency, as confirmed by the results of a computational experiment
- 22 October 2026
- Published by Private Company Technology Center in Eastern-European Journal of Enterprise Technologies
- Vol. 5 (4 (143)) , 58-67
Open AccessDevelopment of a comprehensive method of learning artificial neural networks for organizational and technical systems
The object of the study is the learning process of artificial neural networks (ANNs). The problem addressed in the study is to increase the reliability of ANN learning while ensuring a given level of learning efficiency. The originality of the proposed complex method is: – verification of the architecture of organizational and technical systems (OTS) and the dynamics of its change, due to the use of an improved cheetah pack algorithm. This allows leveling the influence of the human factor and increasing the reliability of ANN learning; – taking into account the possible values of the state of the environment and the actions of agents, which allows taking into account in full the conditions and factors in which ANN learning is carried out. Due to this, the reliability of the ANN learning process increases under the influence of destabilizing factors; – formation of a generalized architecture of multi agent learning due to the use of the ensemble mutation procedure. Due to the use of this procedure, it becomes possible to achieve an increase in the reliability of ANN learning; – selected the type and parameters of the membership function due to the use of the appropriate procedure. Due to the use of this procedure, it becomes possible to achieve an increase in the accuracy of ANN learning; – reduction of sensitivity to a priori hyperparameters of models and tendency to excessive fragmentation of data in OTS, due to the use of improved procedures, which increases the reliability of ANN learning; – for ANN learning, a sample with an uneven distribution of the states of OTS functioning is used, which achieves the flexibility of ANN learning under the influence of destabilizing factors. The proposed complex method of ANN learning minimizes the geometric error of the division by an average of 25%, thereby increasing the reliability of ANN learning while maintaining a given level of efficiency, as confirmed by the results of a computational experiment
- Open Access
- 12 August 2026
- Published by Private Company Technology Center in Eastern-European Journal of Enterprise Technologies
- Vol. 4 (4 (142)) , 39-49
Development of methodology for processing heterogeneous data in special purpose information systems
The object of the study is special-purpose information systems. An issue addressed in the study is the improvement of the efficiency of heterogeneous data processing of special-purpose information systems. A methodology for processing heterogeneous data in special-purpose information systems has been developed, the originality of which is: – comprehensive assessment of the state of processing of heterogeneous data in special-purpose information systems, which allows multi-level management of the process of their processing; – in the ability to conduct unified end-to-end management of the process of processing heterogeneous data in special-purpose information systems; – in the ability to determine optimal management solutions for the processing of heterogeneous data in special-purpose information systems in deterministic conditions by taking into account losses due to non-fulfillment of boundary conditions and restrictions on the inseparable performance of interaction operations; – determination of methods of application of elements and subsystems of heterogeneous data processing of special-purpose information systems due to the calculation of permissible options for the application of elements and subsystems of heterogeneous data processing at the planning interval; – in assessing the stability of operational embodiments and managing elements of special-purpose information systems by taking into account available computing resources. The analysis of the results allows to state that the proposed methodology allows for an increase in the efficiency of heterogeneous data processing by an average of 20%, with the reliability of heterogeneous data processing at the level of 90-95%. At the same time, convergence to local and global optima is ensured, while ensuring an average level of computational complexity
- 12 August 2026
- Published by Private Company Technology Center in Eastern-European Journal of Enterprise Technologies
- Vol. 4 (4 (142)) , 39-49
Open AccessDevelopment of methodology for processing heterogeneous data in special purpose information systems
The object of the study is special-purpose information systems. An issue addressed in the study is the improvement of the efficiency of heterogeneous data processing of special-purpose information systems. A methodology for processing heterogeneous data in special-purpose information systems has been developed, the originality of which is: – comprehensive assessment of the state of processing of heterogeneous data in special-purpose information systems, which allows multi-level management of the process of their processing; – in the ability to conduct unified end-to-end management of the process of processing heterogeneous data in special-purpose information systems; – in the ability to determine optimal management solutions for the processing of heterogeneous data in special-purpose information systems in deterministic conditions by taking into account losses due to non-fulfillment of boundary conditions and restrictions on the inseparable performance of interaction operations; – determination of methods of application of elements and subsystems of heterogeneous data processing of special-purpose information systems due to the calculation of permissible options for the application of elements and subsystems of heterogeneous data processing at the planning interval; – in assessing the stability of operational embodiments and managing elements of special-purpose information systems by taking into account available computing resources. The analysis of the results allows to state that the proposed methodology allows for an increase in the efficiency of heterogeneous data processing by an average of 20%, with the reliability of heterogeneous data processing at the level of 90-95%. At the same time, convergence to local and global optima is ensured, while ensuring an average level of computational complexity
- Open Access
- 30 May 2026
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (105) , 151-158
Using iNav autopilot in a mission simulator for training UAV crews
The article discusses the use of the iNav autopilot as part of a mission simulator for training crews of aircraft-type unmanned aerial vehicles. The subject of the study is the construction of a closed simulation loop in which the autopilot logic works together with a flight dynamics model, virtual sensors, control channels and training scenarios. The aim of the work is to describe the architecture and methodology for integrating iNav in SITL mode for use in a UAV crew simulator. To achieve this goal, the main tasks are defined: to analyze the place of iNav among approaches to autopilot simulation, to describe the order of component launch, to determine the composition of data transmitted between the simulation module, JSBSim and iNav, to consider the role of the MSP and X-Plane UDP protocols, and to form an approach to modeling weather effects, failures and distortions. The research methodology is based on the system modeling of UAV operation in a training environment. JSBSim is used as a source of a dynamic flight model, and the simulation module performs the transformation of coordinates, velocities, angles, accelerations, altitude, airspeed and other parameters into a form suitable for processing by the autopilot. Radio control commands, telemetry, and diagnostic states are transmitted via MSP, and a high-frequency circuit for exchanging sensor values and control outputs is formed via X-Plane UDP. The possibility of transitioning from SITL to HITL, when a physical controller is used instead of a software process, is separately considered. The main results are a generalized architecture of iNav integration with JSBSim, a description of the startup sequence, definition of a set of virtual sensors and output commands, as well as formalization of scenarios of GPS loss, navigation data substitution, electronic jamming, control channel loss, IMU noise, barometer drift, Pitot tube failure, servo saturation, and battery voltage sag. It is concluded that iNav can be effectively used in a fixed-wing UAV simulator provided the simulation bridge is correctly constructed. The practical significance of the work lies in the possibility of creating a safe, repeatable and controllable environment for training crews and practicing actions in normal and abnormal conditions.
- 30 May 2026
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (105) , 151-158
Open AccessUsing iNav autopilot in a mission simulator for training UAV crews
The article discusses the use of the iNav autopilot as part of a mission simulator for training crews of aircraft-type unmanned aerial vehicles. The subject of the study is the construction of a closed simulation loop in which the autopilot logic works together with a flight dynamics model, virtual sensors, control channels and training scenarios. The aim of the work is to describe the architecture and methodology for integrating iNav in SITL mode for use in a UAV crew simulator. To achieve this goal, the main tasks are defined: to analyze the place of iNav among approaches to autopilot simulation, to describe the order of component launch, to determine the composition of data transmitted between the simulation module, JSBSim and iNav, to consider the role of the MSP and X-Plane UDP protocols, and to form an approach to modeling weather effects, failures and distortions. The research methodology is based on the system modeling of UAV operation in a training environment. JSBSim is used as a source of a dynamic flight model, and the simulation module performs the transformation of coordinates, velocities, angles, accelerations, altitude, airspeed and other parameters into a form suitable for processing by the autopilot. Radio control commands, telemetry, and diagnostic states are transmitted via MSP, and a high-frequency circuit for exchanging sensor values and control outputs is formed via X-Plane UDP. The possibility of transitioning from SITL to HITL, when a physical controller is used instead of a software process, is separately considered. The main results are a generalized architecture of iNav integration with JSBSim, a description of the startup sequence, definition of a set of virtual sensors and output commands, as well as formalization of scenarios of GPS loss, navigation data substitution, electronic jamming, control channel loss, IMU noise, barometer drift, Pitot tube failure, servo saturation, and battery voltage sag. It is concluded that iNav can be effectively used in a fixed-wing UAV simulator provided the simulation bridge is correctly constructed. The practical significance of the work lies in the possibility of creating a safe, repeatable and controllable environment for training crews and practicing actions in normal and abnormal conditions.
- 23 December 2025
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (103) , 126-136
Synthesis of a computer information and analytical system for emergencies. Part 1
The paper explores the concept of synthesising a computer information and analytical system for monitoring, analysing, and supporting decision-making during emergencies. The paper includes a review of emergency classifications, an analysis of existing information systems, and a justification for choosing the architecture of the future system. The goal of this work is to develop a computer information and analytical system that enables real-time analysis of the current situation, forecasts the development of events, and provides recommendations for responding to emergencies.
- 23 December 2025
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (103) , 126-136
Synthesis of a computer information and analytical system for emergencies. Part 1
The paper explores the concept of synthesising a computer information and analytical system for monitoring, analysing, and supporting decision-making during emergencies. The paper includes a review of emergency classifications, an analysis of existing information systems, and a justification for choosing the architecture of the future system. The goal of this work is to develop a computer information and analytical system that enables real-time analysis of the current situation, forecasts the development of events, and provides recommendations for responding to emergencies.
- 23 December 2025
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (103) , 137-143
Synthesis of a computer information and analytical system for emergencies. Part 2
The article considers the concept of synthesizing a computer information and analytical system for monitoring, analysis, and decision-making support during emergencies. A functional model of the prototype is developed, a description of the implementation of key system modules is provided, and the use of machine learning methods for assessing the degree of threats is proposed. The work includes an applied software solution and the results of its verification based on open source scenarios. The practical significance lies in creating a prototype of a universal system that can be adapted for the needs of the State Emergency Service of Ukraine, local authorities, or critical infrastructure enterprises.
- 23 December 2025
- Published by Kharkiv National University of Radioelectronics in Bionics of Intelligence
- Vol. 2 (103) , 137-143
Synthesis of a computer information and analytical system for emergencies. Part 2
The article considers the concept of synthesizing a computer information and analytical system for monitoring, analysis, and decision-making support during emergencies. A functional model of the prototype is developed, a description of the implementation of key system modules is provided, and the use of machine learning methods for assessing the degree of threats is proposed. The work includes an applied software solution and the results of its verification based on open source scenarios. The practical significance lies in creating a prototype of a universal system that can be adapted for the needs of the State Emergency Service of Ukraine, local authorities, or critical infrastructure enterprises.