Кваліфікаційні роботи здобувачів вищої освіти кафедри авіаційних комп'ютерно-інтегрованих комплексів (НОВА)

Permanent URI for this collectionhttp://er.nau.edu.ua/handle/NAU/58725

Browse

Now showing 1 - 20 of 167

Multi-agent simulation system of navigation equipment test stand. Modeling of electrical and mechanical parts
(National Aviation University, 2024-06-13) Shykula, Yuliia Yuriivna; Шикула, Юлія Юріївна
A MAS is a system made up of multiple interacting software programs or robots called agents. These agents can be simple or complex. They can sense their environment, can be the physical world or a computer simulation, learn and adapt based on their experiences, can communicate with other agents. Agents in a MAS can communicate, cooperate, coordinate, and compete with each other to accomplish individual or collective tasks. MAS finds applications in various domains such as robotics, distributed systems, economics, traffic management, and social simulation. It allows modeling complex systems where individual components have their own goals, knowledge, and capabilities, leading to emergent behaviors and system-level functionalities that may not be achievable by individual agents alone. MAS typically involves studying agent architectures, communication protocols, coordination mechanisms, negotiation strategies, and decision-making processes. Research in MAS aims to understand how to design, develop, and manage systems composed of multiple autonomous agents to achieve efficient, robust, and adaptive behavior. MAS are used to solve problems that are too complex for a single agent to handle. Each robot could be an agent, with its own sensors to navigate the environment and avoid obstacles, and the ability to communicate with other robots to divvy up the cleaning tasks. MAS offer a powerful and versatile approach to tackling complex problems. Their scalability, flexibility, and robustness make them well-suited for a wide range of applications, from self-driving cars to financial markets. As research in artificial intelligence continues to advance, multi-agent systems are poised to play an increasingly important role in shaping the future.
Mobile robot to search of object with given properties
(National Aviation University, 2024-06-15) Shpachynskyi, Dmitro Oleksandrovich; Шпачинський, Дмитро Олександрович
The purpose of this work is to provide a comprehensive analysis of mobile robot systems for detecting objects with specified properties, focusing on the technical measures of mine clearing and the automation of humanitarian demining processes. The study aims to explore the functional capabilities of mobile robots in law enforcement activities, including emergency situations and environmental safety departments. By investigating the structure, composition, and implementation of subsystems of mobile robots of free orientation, the research seeks to enhance understanding of the complex technological processes involved in mobile robot operations. The object of this work is to examine the opportunities, prospects, and challenges associated with the creation and deployment of mobile robots for detecting objects with specified properties. By analyzing the automation of humanitarian demining processes and the use of remote-controlled and autonomous vehicles in mine detection operations, the study aims to shed light on the advancements in mobile robot technology. Through a detailed exploration of experiment results and conclusions, the research seeks to provide valuable insights into the capabilities and limitations of mobile robot systems for object detection tasks.
Automated system for improving the energy efficiency of maintenance facilities
(National Aviation University, 2024-06-13) Fedorets, Boris Sergiyovych; Федорець, Борис Сергійович
The objective of this thesis is to design, develop, and evaluate an automated system for improving the energy efficiency of maintenance facilities. The subject of this thesis is the development and evaluation of an automated system designed to improve the energy efficiency of maintenance facilities. This encompasses the exploration and implementation of advanced technologies, methodologies, and algorithms aimed at optimizing energy usage and reducing operational costs while maintaining or enhancing operational performance and occupant comfort. The aim of this qualification work is to design, develop, implement, and evaluate an automated system specifically tailored to improve the energy efficiency of maintenance facilities. The research method for this qualification work encompasses a comprehensive approach that includes system design, development, implementation, data collection, analysis, and evaluation. The theoretical research for this qualification work delves into the underlying principles, theories, and models that inform the design and implementation of an automated system aimed at improving energy efficiency in maintenance facilities. Research Results- The automated system for improving energy efficiency in maintenance facilities was successfully developed and implemented. Recommendations Continuously optimize energy use, provide user training, plan for scalability, monitor long-term performance, ensure regulatory compliance, foster collaboration, and invest in research for ongoing innovation.
Improving Mesh Network Efficiency for UAV Control Using Machine Learning
(National Aviation University, 2024-06-15) Smishchuk, Bogdan Mykolayovych; Сміщук, Богдан Миколайович
In recent years, unmanned aerial vehicles (UAVs) have become an integral part of many industries, including agriculture, environmental monitoring, rescue operations, logistics, and military operations. Their ability to perform complex tasks in conditions where human presence may be dangerous or impossible significantly increases the efficiency and safety of operations. However, a reliable communication system capable of adapting to changing environmental conditions is needed to coordinate many UAVs and ensure their smooth operation. Mesh networks are a promising technology for creating such communication systems, as they are characterized by a decentralized structure and the ability to self-organize. These networks are resistant to the failure of individual nodes and provide uninterrupted communication even in difficult conditions. However, to get the most out of mesh networks, the network protocols that control routing and data transmission must be optimized. The application of machine learning techniques offers the potential to optimize network protocols in cellular networks. The application of machine learning algorithms allows analyzing a significant amount of data about the state of the network and its environment in real time. This data is used to automatically adjust network parameters for optimal performance. This leads to improvements in key performance metrics such as latency, throughput and power consumption, which are vital for effective UAV control. The relevance of this topic is due to the need to develop modern technologies that will ensure high reliability and efficiency of UAV control systems. As the popularity and importance of UAVs in various industries continues to grow, optimization of network protocols using machine learning is becoming a critical task. The use of unmanned aerial vehicles in different areas requires the use of different communication and control methodologies. This requires coordination and rapid exchange of information between devices. In the context of rescue operations, the possibility of fast and reliable communication is of paramount importance to save lives. In military operations, the reliability of communication between UAVs is of primary importance for the successful execution of missions. Therefore, research on improving the efficiency of mesh networks for controlling UAVs using machine learning is timely and important, as it meets the modern challenges and needs of the development of unmanned systems. Optimizing communication systems using the latest technologies will significantly increase the efficiency and reliability of UAVs, which will contribute to their wider use and increase the overall level of safety and productivity.
Smart Office Project Based on the use of the Internet of Things Technology
(National Aviation University, 2024-06-15) Sakhno, Oleh Sergiyovych; Сахно, Олег Сергійович
Creating a smart office based on IoT is becoming a necessity in a changing business landscape where productivity and efficiency are becoming key success factors. The introduction of an IoT-based smart office allows companies to meet these requirements, providing them with a competitive advantage in the global market environment. Over the past decades, there has been a rapid development of technologies that transform traditional office spaces into intelligent cosystems. The introduction of IoT allows for the automation, optimization and personalization of work processes, which opens up new horizons for improving work efficiency and employee comfort. Specifically for Kyivstar, the introduction of a smart office is an important step to optimize internal processes, increase employee satisfaction, and improve customer experience. Therefore, studying the capabilities and benefits of IoT in the context of office management is an urgent task that will help increase the company's competitiveness and innovative development. In today's world of increasing competition and rapid technological development, the office environment is being transformed by the application of Internet of Things (IoT) technology. "An IoT-based smart office is defined by a wide range of opportunities to optimize business processes, increase productivity, and improve the quality of life for employees. Recent studies in the field of smart office confirm its potential and benefits. For example, a study by Deloitte found that 75% of companies that have implemented IoT in their offices have seen an improvement in employee productivity. Such technologies allow automating routine tasks, simplifying communication, and providing quick access to data.
Intelligent System for Hyperspectral Image Processing Based on Recurrent Neural Networks
(National Aviation University, 2024-06-14) Sarin, Oleksandr Evgenievich; Сарін, Олександр Євгенійович
Hyperspectral technologies are opening a new window on the world, giving us the ability to see and understand reality in much greater detail than ever before. These technologies allow us to collect information about objects and environments not only in visible light, but also in a wide range of wavelengths, including infrared and ultraviolet. This gives us access to information that is invisible to the human eye and allows us to better understand the chemical composition, structure and physical properties of objects. Hyperspectral technologies are used in many fields, including remote sensing: hyperspectral satellites and aircraft are used to study the earth's surface, monitor the environment, manage land resources and detect natural disasters. In medical diagnostics, hyperspectral images are used to diagnose cancer, cardiovascular disease and other diseases. In food quality control, hyperspectral data is used to detect contaminants in food and assess its chemical composition. In forensic science, hyperspectral data is used to analyse crime evidence such as blood, hair and drugs. In agriculture, hyperspectral imagery is used to assess plant health, detect pests and diseases, and optimise irrigation and fertilisation. In the pharmaceutical industry, hyperspectral data is used to control the quality of medicines and to detect counterfeiting. In art and restoration, hyperspectral images are used to analyse works of art, reveal hidden details and restore damaged paintings.
Automated system for preparing printed circuit boards based on a laser engraver
(National Aviation University, 2024-06-14) Salmi, Mykola Olehovych; Сальмі, Микола Олегович
In today's world of electronics and technology, printed circuit boards play a key role in the functioning of various devices. From simple household appliances to complex industrial systems – printed circuit boards are the basis for placing and connecting electronic components. With the growing demand for electronic devices and increasing their functionality, there is a need for fast and efficient production of printed circuit boards. Traditional PCB manufacturing methods, such as chemical etching and mechanical milling, have certain drawbacks. Chemical etching requires the use of hazardous substances, which can have a negative impact on the environment and human health. Mechanical milling, in turn, has limitations regarding the complexity and accuracy of printed circuit boards. Therefore, it is urgent to search for alternative methods of manufacturing printed circuit boards that would ensure high accuracy, speed and environmental friendliness of the process. One of the promising directions in the production of printed circuit boards is the use of laser technologies. Laser engraving allows you to achieve high accuracy and speed of manufacturing printed circuit boards, while reducing the impact on the environment. Automation of the process of preparing printed circuit boards on the basis of a laser engraver allows to significantly increase the efficiency and productivity of production.
Multi-agent simulation system of navigation equipment test bench. Simulation of control system software
(National Aviation University, 2024-06-13) Rykov, Stanislav Vyacheslavovich; Риков, Станіслав Вячеславович
The primary object of development is the multi-agent simulation system itself, aimed at enhancing the understanding and optimization of navigation equipment behavior under diverse operational conditions. The purpose of this development is to provide a robust tool for researchers and engineers to simulate, analyze, and predict the performance of navigation systems, thereby aiding in design and operational improvements. The development methods employed include the use of Proportional-Integral-Derivative (PID) control algorithms for system stabilization, numerical integration techniques for dynamic modeling, and artificial intelligence algorithms for agent-based decision-making. These methods ensure a high fidelity simulation environment capable of reflecting real-world dynamics. The results of this work demonstrate the system’s capability to accurately simulate complex multi-agent interactions and its adaptability to various types of navigation equipment. The novelty of the system lies in its modular design, which supports scalability and the integration of new technologies or algorithms as they emerge. This adaptability makes the system a forward-looking tool that can evolve with advancements in navigation technology.
Intelligent data processing system
(National Aviation University, 2024-06-14) Petrov, Denis Mykolayovych; Петров, Денис Миколайович
The paper explores and demonstrates efficient methods for time series analysis and forecasting using Python, focusing on weather data from 2013 to 2017. Qualification work will include data preprocessing techniques such as normalization and denoising, and will use Fourier transforms and gradient descent for curve fitting. The goal is to identify patterns and predict future data points using machine learning models built with TensorFlow and evaluated using visualization and error metrics. Weather forecasting is critical to various industries, but forecasting weather conditions remains a challenging task due to the complexity and variability of climate data. This task shows how this problem can be solved using modern methods of time series analysis and machine learning. The focus is on processing weather data from 2013 to 2017 to identify patterns and develop forecasting models. The problem lies in the efficient processing of noisy data, the application of appropriate normalization and transformation methods, and the selection of reliable algorithms to increase the accuracy and reliability of predictions.
Electric energy conversion control system
(National Aviation University, 2024-06-14) Petryna, Вадим Андрійович; Петрина, Vadym Andriyovych
In the work, a study of the applied methods of the theory of electric circuits for the analysis of the processes of the conversion of electrical energy and the principles of automated control to determine the strategy of controlling the power switches of the converter is carried out. The qualification work is dedicated to the promising direction of automation of energy supply systems with the combined use of wind energy, diesel power plants and centralized electrical networks with the help of autonomous energy systems. Improvement of energy conversion processes is one of the directions of increasing the efficiency of electric energy generation in autonomous energy systems. Direct frequency conversion is one of the most promising areas of development of power converters of autonomous power plants. Currently, the use of matrix converters characterized by a high efficiency due to direct single-stage energy conversion in electrical energy conversion systems is of great interest; relatively simple power part; the possibility of energy recovery to the grid; high power factor due to the absence of reactive elements.
An intelligent mobile search system
(National Aviation University, 2024-06-15) Koval, Maksym Anatoliyovych; Коваль, Максим Анатолійович
The field of Synthetic Aperture Radar (SAR) data processing has experienced significant advancements in recent years, completely transforming the way we collect and interpret information for a wide range of applications. SAR data, obtained through radar sensors, plays a crucial role in environmental monitoring, disaster management, urban planning, and various other domains. The exceptional capabilities of SAR technology, including all-weather imaging and high-resolution data collection, have made it an invaluable tool for extracting valuable insights from remote sensing data. We explore the intricate process of SAR data collection, storage, processing, and interpretation, shedding light on the complexities and challenges associated with harnessing the full potential of SAR observations. By gaining a deep understanding of the underlying methodologies and technologies that drive SAR data processing systems, we can truly appreciate the transformative impact of SAR technology on diverse fields. The chapter takes us through the evolution of SAR data processing systems, emphasizing the role of machine learning algorithms, artificial intelligence techniques, and user interfaces in enhancing the efficiency and accuracy of data analysis. We examine the critical role of anomaly detection in environmental monitoring and disaster response, highlighting the significance of early detection and mitigation of potential threats through advanced SAR data processing methods.
Control system of the MRPD–05 microprocessor relay protection device
(National Aviation University, 2024-06-14) Kostiuchenko, Maksym Volodymyrovych; Костюченко, Максим Володимирович
The work has studied and analyzed the principle of operation and the architecture of the control system of the MPZS-05 microprocessor relay protection device for increasing the reliability and efficiency of electrical network protection. Theoretical studies consisted of an in-depth analysis of the architecture and algorithms of microprocessor relay protection devices, as well as studying the principles of design and implementation of control systems. The research results showed that the developed control system of the MPZS-05 microprocessor relay protection device significantly increases the efficiency of responding to emergency situations, reduces the number and duration of emergency shutdowns, and also increases the reliability of power supply. The results can be used to familiarize with the control systems of microprocessor relay protection devices, their features and useful tools.
An intelligent system for detection and classification of anomalous objects
(National Aviation University, 2024-06-14) Kulakevych, Illia Vitaliyovych; Кулакевич, Ілля Віталійович
The current state of the art in security and environmental protection technologies requires continuous improvement of methods for detecting and classifying anomalous objects. Anomalies, such as environmental pollution, oil spills, illegal logging, or potential explosive devices, can have serious consequences for the environment and public safety. Traditional methods, such as metal detectors and X-ray scanners, have limited effectiveness in difficult environments. For example, they are often unable to detect sophisticated explosive devices that can be made of non-metallic materials or disguised as ordinary objects. Also, these methods are often ineffective in detecting environmental anomalies, such as oil spills or water pollution. One of the most promising areas in this field is the use of hyperspectral cameras, which allow analyzing objects in a wide range of electromagnetic radiation. Hyperspectral images contain much more information than traditional color or monochrome images, which opens up new opportunities for identifying and classifying materials based on their spectral characteristics. This is especially important for tasks requiring high identification accuracy, such as detecting environmental contamination or security threats. The relevance of this topic is due to the need to improve the efficiency of methods for detecting anomalies that are critical to environmental safety and public security. To date, research in this area has shown that the use of hyperspectral imagery can significantly improve the accuracy of anomaly identification. However, existing methods of hyperspectral image processing require further improvement, in particular in the direction of automating data analysis and using modern machine learning algorithms. The innovation of this work is the development of an intelligent system for detecting and classifying anomalous objects based on hyperspectral images, which will use advanced neural network methods for data processing and analysis. This will increase the speed and accuracy of detecting potentially hazardous objects, which is extremely important for preventing environmental disasters, terrorist attacks, and ensuring public safety. The aim of this work is to develop an intelligent system for detecting and classifying anomalous objects based on hyperspectral images.
An unmanned aerial vehicle for searching for victims in emergency situations
(National Aviation University, 2024-06-13) Liashenko, Volodymyr Oleksandrovych; Ляшенко, Володимир Олександрович
The main objective of this work is to analyze the efficiency of using unmanned aviation systems in search and rescue operations. Presently, Ukraine encounters a significant number of emergency situations leading to a high number of casualties and fatalities, underscoring the necessity and relevance of search and rescue operations. This type of work is pivotal and falls within the realm of emergency response efforts. Additional measures are devised each year to enhance safety, conduct briefings at potentially dangerous sites, and patrol high-risk areas. However, emergency situations often occur uncontrollably, impacting the number of casualties and affected individuals. Search and rescue operations represent some of the most critical activities conducted by rescue services globally. The primary objective of such operations is to locate, rescue, and provide emergency medical assistance to casualties. Search and rescue teams carry out these operations using both ground and aviation equipment. The advancement of modern technologies significantly enhances efficiency and optimizes the process of searching for and rescuing casualties at all stages. One modern solution is the use of unmanned aerial vehicles during search and rescue operations. The utilization of unmanned aerial vehicles, instead of manned aircraft, shows promise and effectiveness. Unmanned aviation systems are a modern development that remains insufficiently researched. However, in recent years, the application of unmanned aviation systems has rapidly expanded in both civilian and military aviation sectors.
Automated microclimate control system for the passenger cabin of an aircraft
(National Aviation University, 2024-06-13) Klymenko, Oleksandr Andriyovych; Клименко, Олександр Андрійович
In an era where air travel has become an indispensable mode of transportation, ensuring the comfort and well-being of passengers during flights is a critical priority for airlines and aviation authorities worldwide. Central to this endeavor is the management of the microclimate within the aircraft cabin, which encompasses a multitude of environmental factors that directly impact passengers' comfort and overall experience. The concept of microclimate in the aircraft cabin refers to the localized atmospheric conditions that passengers encounter during their journey. It encompasses a dynamic interplay of variables, including temperature, humidity, airflow patterns, air quality, and radiation exposure, all of which must be carefully controlled to create a conducive environment for passengers. Passenger comfort is not merely a luxury but a fundamental aspect of aviation safety and efficiency. Discomfort or distress experienced by passengers can lead to reduced cognitive function, compromised decision-making abilities, and even physiological stress responses, all of which may pose risks to the safety of the flight. Therefore, effective microclimate control systems are essential to maintain optimal conditions within the cabin and ensure the well-being of passengers throughout their journey. Furthermore, the significance of microclimate control extends beyond passenger comfort to encompass broader considerations such as health, hygiene, and disease prevention. In the context of the ongoing COVID-19 pandemic, the importance of maintaining clean and healthy cabin environments has been underscored, with enhanced ventilation, air filtration, and sanitation measures becoming critical components of aviation safety protocols.
An intelligent system for predicting the need to replace aircraft parts based on transformers
(National Aviation University, 2024-06-15) Dziuba, Andrii Olegovych; Дзюба, Андрій Олегович
This study centers on crafting and implementing an intelligent framework for anticipating aviation part replacements through transformer models. Transformers, a sophisticated class of deep learning models celebrated for their exceptional performance in natural language processing, have recently demonstrated significant potential in diverse predictive assignments. Leveraging transformers, this study strives to establish an advanced predictive maintenance system adept at precisely forecasting optimal part replacement timings. The proposed system is tailored to scrutinize extensive operational and sensor data from aircraft, identifying pre-failure patterns and irregularities. By foreseeing part replacement probabilities and schedules, the system can aid airlines and maintenance crews in refining their maintenance calendars, curtailing downtime, and curbing expenses. Furthermore, this methodology aligns with the industry's progression towards predictive maintenance, emphasizing data-centric decision-making to enhance operational efficiency and safety.
The system of remote control of the vertical movement of the aircraft
(National Aviation University, 2024-06-15) Mykhailenko, Danylo Igorovich; Михайленко, Данило Ігорович
The object of research is a system of remote control of the vertical movement of an aircraft. The subject of the study is a multi-purpose UAV (quadrocopter). The purpose of the qualification work is to develop the software code for the remote control of the vertical movement of the UAV, and the calculation. The task of the qualification project is the development of a remote control system for the movement of UAVs, taking into account calculations of flight characteristics. Theoretical studies consisted of analysis of existing systems, development of a remote control system algorithm, development of software code for remote control of an aircraft in the vertical plane. The results of the qualification work are recommended for familiarization with remote control in the vertical plane, the program code of the remote control of the aircraft, the mathematical model of the UAV, the kinematic model. For more accurate calculations, it is recommended to add a dynamic mathematical model to the system for dynamic motion calculations.
Language command system in on-board control systems
(National Aviation University, 2024-06-13) Didyk, Viacheslav Andriyodych; Дідик, Вячеслав Андрійович
The integration of language command systems within on-board control systems stands at the forefront of aviation innovation, representing a pivotal leap forward in aircraft technology. In an era characterized by ever-increasing automation and complexity in aviation operations, the implementation of advanced language interfaces holds immense promise for revolutionizing the way aircraft are controlled and managed. The journey towards integrating language command systems into aircraft control spans decades, marked by significant milestones and breakthroughs in speech recognition, natural language processing, and human-machine interaction. Understanding the historical context provides crucial insights into the evolution of these systems, tracing their development from rudimentary prototypes to sophisticated, real-world applications in modern aviation. At the heart of language command systems lie a myriad of cutting-edge technologies, including advanced speech recognition algorithms, machine learning models, and natural language understanding frameworks. This work delves into the intricate technical aspects behind the functioning of these systems, unraveling the complexities of speech processing, pattern recognition, and semantic analysis that enable seamless interaction between pilots and aircraft systems.
Aircraft fuel measuring system based on pressure sensors
(National Aviation University, 2024-06-13) Duka, Yrii Serhiyovych; Дука, Юрій Сергійович
The use of a hydrostatic sensor in the fuel measuring system during maneuvering of the aircraft made it possible to achieve more accurate and reliable measurements in contrast to other methods of estimating the remaining fuel. The developed system allows you to effectively take into account the influence of various factors, such as pressure, temperature, flight altitude, changes in the position of the aircraft during movement on the determination of the remaining fuel. The results of the work can be used in the practical activities of airlines to optimize the fuel measuring system and improve flight safety. Further research can be aimed at improving technical solutions and expanding the capabilities of the system in different operating conditions of the aircraft.
Quadruped robot walking algorithm
(National Aviation University, 2024-06-13) Innokov, Oleksandr Olegovich; Інноков, Олександр Олегович
This work develops a robot observer that can help rescuers speed up their work when searching for victims in a mountainous area. In recent years, there has been a great deal of research on robots, especially robots with legs, including bipedal, tripedal, quadrupedal and six-legged robots. The mechanical design of multi-legged robots certainly has advantages and disadvantages compared to wheeled robots. Multi-legged robots can successfully traverse uneven surfaces such as rocks, but they are not faster than wheeled robots. Of course, more leg configurations create their own problems, such as synchronizing the movement of each robot's legs to the desired position and coordinating their motion patterns. Mobile robots with legs as a locomotion system have been the subject of much research. But at the same time, legs have a number of problems. In fact, research on legged robots focuses on everything related to the movement and coordination of the legs during the robot's navigation. The main challenge in developing a legged robot is the coordination of the legs' degrees of freedom. It is the coordination of a large number of degrees of freedom that must be triggered in a simple movement of the robot from point A to point B. Precise coordination of walking movements, the precise coordination of the walking motion is to ensure that the dynamic forces generated during navigation keep the robot in balance. This is a major challenge. These limitations can be overcome by modeling statically. The dynamic balance of forces due to the robot's movements can be overcome by statically modeling it.

Browse

Recent Submissions