2024
Постійне посилання на фонд
Переглянути
Перегляд 2024 за Назва
Зараз показуємо 1 - 20 з 31
Результатів на сторінці
Налаштування сортування
Документ Відкритий доступ Advanced Edge Detection Techniques for Enhanced Diabetic Retinopathy Diagnosis Using Machine Learning(КПІ ім. Ігоря Сікорського, 2024) Basarab, M. R.; Ivanko, K. O.Diabetic retinopathy (DR) represents one of the most serious complications associated with diabetes mellitus, posing a significant threat to vision and leading to severe impairment and potential blindness if not diagnosed and treated promptly. The study investigates the integration of advanced edge detection techniques with machine learning algorithms to enhance the precision and effectiveness of DR diagnosis. By leveraging the APTOS 2019 Blindness Detection dataset, the research employs a combination of edge detection methods such as the Sobel operator and the Canny edge detector, alongside advanced preprocessing techniques and sophisticated feature extraction methods. The study reveals that the synergy between these edge detection techniques and machine learning significantly boosts the diagnostic accuracy of neural networks. Specifically, the accuracy for multiclass classification (spanning five categories: No diabetic retinopathy, Mild, Moderate, Severe, and Proliferative diabetic retinopathy) improved from 78.5% to an impressive 88.2%. This marked enhancement underscores the potential of these techniques in refining the diagnostic processes for early DR detection. By improving the accuracy of classification, this approach not only facilitates early intervention but also plays a crucial role in reducing the risk of severe vision loss among patients with diabetes. The findings of this study emphasize the importance of integrating advanced image processing techniques with machine learning frameworks in medical diagnostics. The improved outcomes demonstrated in this research highlight the potential for such technological advancements to contribute meaningfully to the field of ophthalmology, leading to better patient care and potentially transforming the standard of practice in DR diagnosis.Документ Відкритий доступ Classification of Structural and Functional Development Stage of Cardiomyocytes Using Machine Learning Techniques(КПІ ім. Ігоря Сікорського, 2024) Bondarev, V. R.; Ivanko, K. O.; Ivanushkina, N. G.The study is dedicated to the problem of classification of structural and functional development stage of cardiomyocytes derived from the induced pluripotent stem cells with application of the digital image processing methods and machine learning algorithms, in particular, neural networks. Cell regenerative therapy has become one of the most promising treatment options for patients with heart failure. But since cardiomyocytes are objects with a high level of complexity and have significant morphological variability, automatic classification is complicated by the lack of implemented methods. That's why researches in this area are a major global public health priority. The initial data set used in this study is a publicly open set of confocal microscopic images of cardiomyocytes which can be divided into five classes based on the morphological features (the structure of the transverse T-tubule). A small amount of input data leads to the need of using augmentation methods. Methods that prevent the alteration of the transverse T-tubule, which is an important parameter for correct classification of the development of cardiomyocytes, are used. Histogram equalization technique is used to enhance the contrast and dynamic range of the confocal microscopic images with the method of contrast-limited adaptive equalization. This helped to improve the local contrast of the analyzed images and highlight the main elements of the cardiomyocyte. Finally, Chan–Vese method, which belongs to the regional segmentation methods, is chosen for the image segmentation and removing artifacts and/or parts of other cells from the image. A pre-processed and augmented dataset is used for training of the convolutional neural network having an architecture with hierarchical structure and residual block usage. The model is evaluated based on the confusion matrix and the heat maps of different convolutional layers are analyzed. Images from the classes with a large number of mutual errors are also considered. Based on the analysis, several classes of structural and functional development of cardiomyocytes are combined. Final accuracy of the model for defining the cardiomyocytes maturation stage achieved 77%.Документ Відкритий доступ Designing Minimalistic Powered Arm Orthosis for Brachial Plexus Injuries(КПІ ім. Ігоря Сікорського, 2024) Kotsiubailo, A. V.; Savchuk, A. V.; Omelkyna, D. V.; Shoferystov, S. Ye.; Lavrenko, Ia. I.; Tretiak, I. B.; Yakovenko, S. M.; Lysenko, O. M.; Popov, A. O.Elbow paresis, often resulting from brachial plexus injury, presents a significant challenge in the field of rehabilitation. To address this, we have developed a prototype powered orthosis that utilizes non-invasive surface electromyography (EMG) signals from neck muscles, such as the sternocleidomastoid, for intuitive control. This EMG-driven system allows for the precise manipulation of the elbow joint, covering the full physiological range of motion. The prototype’s design integrates an EMG signal processor with an orthosis action operator, creating a seamless interface between human intent and mechanical action. Healthy participants were able to use neck muscle contractions to control elbow rotation effectively, demonstrating the system’s potential for real-world application. The scaled EMG envelope directly influences the orthosis’s rotational actuator, ensuring responsive and accurate control. Through rigorous sensitivity analysis, we optimized the control algorithm by adjusting EMG window lengths, signal filtering, and thresholding parameters. This optimization process ensures that the system can adapt to individual user needs, providing personalized and efficient control. The real-time control achieved with this prototype marks a significant step forward in the development of biomedical rehabilitation devices. It not only offers a practical solution for those affected by elbow paresis but also lays the groundwork for future advancements in neuromechanical interfaces. Our ongoing research aims to refine this technology further, exploring the integration of signal processing algorithms to predict and adapt to user movements, thereby creating a more natural and intuitive user experience. The ultimate goal is to develop a fully functional orthosis that can be readily implemented in clinical settings, providing a non-invasive, effective solution for elbow rehabilitation.Документ Відкритий доступ Impact of Electronic Components Thermal Resilience on the Reliability of Radio-Electronic Equipment(КПІ ім. Ігоря Сікорського, 2024) Nikitchuk, A. V.The article focuses on the impact of temperature on the reliability of electronic components, as in non-redundant radio-electronic equipment, the failure of any component typically leads to the failure of the entire device. The methods and approaches used to analyze the electronics reliability, predict operational lifespans, and to enhance it are considered. Thermal effects are among the most significant factors influencing reliability indicators of electronics, such as the probability of failure-free operation and mean time to failure. The sequence of accounting for thermal factors during the calculations of operational failure rate, mean time to failure, and the probability of failure-free operation according to the recommendations of Ukrainian State Standards is analyzed. The primary focus is on calculating the mean time to failure for various groups of resistors, capacitors, integrated circuits, and semiconductor components. Modern approaches to reliability assessment are used in the study, particularly a combination of failure physics and computer modeling. It was determined that the difference in the mean time to failure between the most and least thermally resilient electronic components of radio-electronic equipment can be very significant and only increases with rising temperatures.Документ Відкритий доступ Impact of Gamma, Neutron, Ion, and Electron Irradiation on Structure and Properties of Graphene(КПІ ім. Ігоря Сікорського, 2024) Biliak, R. V.This article discusses studies of the effects of various types of radiation, including y-radiation, neutron, ion, and electron irradiation, on graphene and graphene-based devices. The study of graphene’s response to radiation is crucial because of its potential applications in fields such as nuclear power and space exploration, where the impact of radiation is significant. This paper discusses recent experiments conducted to investigate the effects of y-radiation on graphene layers and devices based on graphene, which revealed changes in graphene layer spacing, defect formation, and electrical characteristics. Similarly, studies of the effect of neutron irradiation on graphene demonstrate its resistance to such radiation, with graphenebased sensors retaining functionality even after exposure to high neutron flux. Moreover, studies of ion irradiation reveal the ability to modify the structure of graphene, although it causes significant damage. The electron irradiation creates defects, which in turn reduced noise levels in graphene-based devices, a unique characteristic not observed in traditional materials. Conclusions were also drawn regarding the effect of temperature on graphene, and it was found that elevated temperatures contribute to the reduction of defects through annealing, demonstrating the self-healing properties of graphene. The paper concludes by emphasizing graphene’s resistance to radiation and its potential for use in high radiation environments where traditional materials may not be able to withstand. The results show that graphene-based sensors and devices can maintain functionality even in the presence of defects caused by radiation, which opens up promising prospects for applications in nuclear energy and space research. Further in-situ studies are recommended to better understand the real-time effects of radiation on device functionality. Overall, graphene is an excellent candidate for various applications due to its unique properties and radiation resistance.Документ Відкритий доступ Implementation of Method of Minimizing the Side Lobe Level of Autocorrelation Functions of Signals With Nonlinear Frequency Modulation(КПІ ім. Ігоря Сікорського, 2024) Kostyria, O. O.; Hryzo, A. A.; Khizhnyak, I. A.; Fedorov, A. V.; Lukianchykov, A. A.An important problem that is solved during the creation of new and modernization of existing radar equipment is to ensure the maximum range of detection of air targets, which requires increasing the radiated power while maintaining the required range resolution. Since the generating devices, which are now widely used as semiconductor elements, have limited peak power, the required energy is emitted by increasing the duration of the sensing radio pulse, and the requirements for resolution are met by using so-called complex signals, the product of the spectrum width of which and their duration (signal base) is greater than one. One of the types of complex signals is multifragment signals with nonlinear frequency modulation, which, unlike the well-known linear frequency modulated signals, have a significantly lower peak side lobes level of autocorrelation functions, but the value of this level depends significantly on the frequency and time parameters of the signal. Finding parameters that minimize the side lobes level of the autocorrelation function of nonlinear frequency modulated signals, which include fragments with linear frequency modulation, is an important scientific and technical problem, the solution of which is the subject of this article. The peculiarity of considering this issue is that, in contrast to the previously proposed implementation of the method of minimizing the side lobes level for mathematical models with a current time change, the paper develops models of shifted time, that is, when the time count of each subsequent signal fragment is shifted to zero. The first section of the paper analyzes the known publications, which shows that the method of minimizing the side lobes level of correlation functions has not been considered before for mathematical models of the shifted. Given this circumstance, the second section of the paper formulates the research objectives. The theoretical justification of a new variant of the proposed method by developing mathematical time-shifted models for two- and three-fragment nonlinear frequency-modulated signals, as well as the modeling results, are presented in Section 3. In further research, it is planned to develop an algorithm for optimizing the time-frequency parameters of signals with nonlinear frequency modulation based on mathematical models of current and shifted time.Документ Відкритий доступ Investigation of Possibility of Measuring the Albedo of Earth’s Surface in Visible and Near Infrared Bands in Conditions of Aerosol Pollution of Atmosphere Using Unmanned Aerial Vehicles(КПІ ім. Ігоря Сікорського, 2024) Asadov, H. H.; Alieva, A. J.; Ashrafov, M. G.It is well-known that such processes as agricultural activities, urbanization processes, climatic changes leading to abnormal precipitation, etc. affect the magnitude of the Earth’s albedo. At the same time, the results of remote albedo measurement in visible and near infrared bands also depend on the degree of aerosol pollution of the atmosphere. These factors leading to changes in the earth’s albedo lead to the need for periodic measurements of regional values of the albedo of the earth’s surface. There are a number of problems related to measuring the albedo of the earth, related to the spatial and temporal variability of this indicator. These include the dependence of albedo on the zenith angle of the Sun, the need to create albedo measurement networks in the form of numerous geographically distributed pyranometers, the dependence of satellite albedo measurements on the state of the atmosphere, leading to the need for inter-satellite calibration, or groundbased validation measurements. At the same time, the issue of fully accounting for the effect of atmospheric aerosol on the results of measuring the albedo of the Earth’s surface is still open. The article is devoted to the measurement of the Earth’s albedo visible and near infrared bands using UAVs in conditions of aerosol pollution of the atmosphere. The model of single scattering of the optical source signal of an atmospheric aerosol was adopted as the basis of the conducted research. The interrelation of such optical indicators as the optical thickness of the aerosol and the albedo of the Earth’s surface is analyzed. A criterion for the effectiveness of atmospheric measurements using UAVs is proposed, in which efficiency is defined as the ratio of the total radiation entering the on-board spectroradiometer to the amount of extra-atmospheric radiation from the Sun. By switching from a discrete model to a continuous model created to calculate the proposed efficiency criterion, it is shown that with a synchronous change in the optical thickness of the aerosol and albedo, according to the calculated law, the minimum efficiency of measurements of the albedo of the earth’s surface is achieved.Документ Відкритий доступ Mathematical Model of Shifted Time of Combined Signal as Part of Fragments with Linear and Quadratic Frequency Modulation(КПІ ім. Ігоря Сікорського, 2024) Kostyria, O. O.; Hryzo, A. A.; Solomonenko, Yu. S.; Dodukh, O. M.; Biernik, Ye. V.The issues of synthesis of nonlinear-frequency modulated probing signals, which, in comparison with wellknown linear-frequency modulated signals, have a lower maximum level of side lobes of the autocorrelation function, have a practical orientation, relate to actual problems of theory and practice of forming signals with intra-pulse modulation for radio electronic means for various purposes. The paper considers a combined nonlinear-frequency modulated signal consisting of linearly and quadratically-frequency modulated fragments. The peculiarity of the proposed approach to the description of its mathematical model is the introduction of frequency-phase compensation components, which reduces the maximum level of the side lobes of the autocorrelation function of the signal. Calculation of values of compensation components is based on consideration of the effect of derivatives of function of instantaneous phase of fragments up to the most significant order inclusive. The limitation of the method should include the requirement for the existence of their final quantity. In the first section of the article, an analysis of known studies and publications is carried out, which states that for the mathematical model of shifted time, considered in the work, the proposed method of compensating for frequency-phase distortions has not been previously considered. Therefore, in the second section of the work, the corresponding task of the study is formulated. In order to achieve the formulated task of research, the third section of the work develops a mathematical model of the shifted time of the combined signal, which contains the compensatory components of the specified distortions. The importance of their consideration in the resulting signal is theoretically substantiated and clearly demonstrated. As a result of the research, the theory of synthesis of combined signals has been developed, the composition and determination of the magnitude of frequency-phase distortions, which are caused by the appearance of the third derivative of the instantaneous phase function of the quadraticallyfrequency modulated fragment, have been established. As a direction for further research, it is planned to develop and study a mathematical model of shifted time of a three-fragment combined signal with a quadratically-frequency modulated fragment.Документ Відкритий доступ Mathematical Modeling the Electrical Impedance of Piezoceramic Disk Oscillating in Wide Frequency Range (Part 2. Medium Frequencies)(КПІ ім. Ігоря Сікорського, 2024) Bazilo, C. V.; Faure, E. V.; Usyk, L. M.; Tuz, V. V.; Chorniy, A. M.This paper presents further developments in mathematical modeling of the electrical impedance of a piezoceramic disk in a wide frequency range, specifically focusing on the mid-frequency range, i.e., when the elastic wavelength becomes commensurate with the radius of the piezoceramic disk, which is important for numerous modern applications. A mathematical model was developed for disk piezoelectric transducers made of piezoceramics to estimate their electrical impedance and quasi-static electrical capacity in the medium frequency range basing on their geometrical, physical, and mechanical characteristics. The research has found that a piezoceramic disc attains electromechanical anti-resonance in the medium frequency range at frequency, at which its electrical impedance follows to infinity. This effect is due to the polarization charges being completely compensated by the electric charge, when the electric current vanishes and energy consumption from the generator is absent. The calculations proved that at frequencies close to the first thickness resonance (corresponding to the dimensionless wave number from 40 to 60), the radial displacements of material particles of the disk vanish. A very rapid decrease in the levels of radial shifts with a simultaneous increase in the electromechanical resonance number was noted. The evaluation of the mechanical quality factor of piezoceramic disk elements, obtained with the developed mathematical model, closely correlates with real values, which is confirmed by the high agreement between theoretical and experimental results.Документ Відкритий доступ Method for Detecting Small Aerial Objects Appearing in Field of View in Controlled Part of Celestial Sphere in Infrared Range(КПІ ім. Ігоря Сікорського, 2024) Agaev, F. G.; Asadov, H. H.; Aliyeva, G. V.The article is devoted to the developed method of infrared detection of group remote high-temperature objects. The problem of searching for the extremum of the total infrared radiation of a group of non-identical thermal objects carrying out a group flight is formulated and solved using the variational optimization method. Examples of such objects include the flight of aircraft in a group, ground scenes involving a group of objects of interest, temperature diagnostics of various points of buildings, control of automobile traffic on highways, control of group flights of birds, drones, etc. A condition has been determined under which the total value of the infrared radiation flux of thermal elements in the group reaches an extreme value. The regression relationship function between the emission coefficient of the thermal elements of the group and the atmospheric transmission coefficient has been calculated. The problem of optimal control of small thermal objects randomly distributed in the atmosphere is practically solved using a ground-based multiradar system in which elements of a multi-radar system monitor flying objects with different values of the radiation coefficient on the routes and different atmospheric transparencies. The proposed method can be used for remote control of flight or the functioning of a group of flying thermal objects with different values of the radiation coefficient with a special procedure for selecting a controlled aircraft for observation by an element of a multi-radar system. The property of the extremum of the total IR radiation flux was found in the inverse relationship between the radiation coefficients of all controlled flying objects and the transparency of the atmosphere along the route between the multi-radar element and the controlled flying object.Документ Відкритий доступ Method for Selecting Pulsed Signals by Their Duration in Fading Channels(КПІ ім. Ігоря Сікорського, 2024) Buhaiov, M. V.; Zakirov, S. V.The duration of pulsed signals is one of the main parameters to be estimated in radio monitoring systems. When signals propagate in channels with deep fading, even at high signal-to-noise ratios, the pulse shape will be distorted. In sophisticated electronic environment, it is also may be random interference in signal processing channel, which leads to the occurrence of false pulses with random durations. Therefore, the values of the signal pulses durations will be concentrated near their true value, and the rest of the detected pulses will have a significantly random duration. That’s why, the development and study of methods for selecting pulse signals by their durations in sophisticated signal environment is actual scientific problem. The aim of the work is improving pulsed signals processing methods in fading channels by selecting its’ durations. The study found that the estimates of signal pulse durations are normally distributed. Pulse durations that are not related to signals are subjected to an exponential distribution. The input data for the proposed method is only a sample of measured pulse durations. The values of the parameters of both the exponential and normal distributions are unknown. In this case, the problem of selecting pulses by their durations is formalized to the estimation of the mean values of normal distributions. To do this, it is proposed to search for the maxima of the smoothed estimate of the probability density function. The scientific novelty of the obtained results is that a method for estimating the mean value of a normal distribution at the background of exponentially distributed values was proposed. An example of this approach is the estimation of pulsed signal durations in channels with deep fading and impulse interference. Based on the developed method, algorithms for automatic pulse selection for radio monitoring systems can be implemented.Документ Відкритий доступ Neural Networks Detection of Low-Amplitude Components on ECG Using Modified Wavelet Transform(КПІ ім. Ігоря Сікорського, 2024) Mnevets, A. V.; Ivanushkina, N. G.This study is devoted to identification of low amplitude components from ECG signals by different timefrequency analysis methods when main power spectrum falls on high-amplitude components. It was also analyzed the problem of choosing correct scale system for determination low-amplitude components on the scalogram by artificial intelligence models. As a result of the study, several modifications of the continuous wavelet transform were proposed. First modification was based on the use of a scaling function and a modified wavelet. Second modification was based on the use of cosine similarity at each iteration of convolution followed by the use of a scaling function. The main idea of the study was to modify the wavelet transform in such a way as to select the components which has the target amplitude and reduce all other components that complicate the neural networks analysis of the interested fragments of the signal. Also, possible procedures for signal restoring were proposed for preserving the effect of using scaling modifications. The testing of the proposed modified algorithms was carried out on the basis of artificially created signals as well as on the basis of real ECG signals with late potentials superimposed on them. Visual analysis of scalograms and signal reconstructions obtained using the modified wavelet transform showed that the modified wavelet transform is capable of extracting low-amplitude components from the signal with much greater spectral power than the transform without modifications. In addition, the ability of common neural network models to distinguish between cardiac cycles with and without late potentials was tested. As a result, it was found that models that were trained on scalograms obtained using a modified wavelet transform train faster and are less susceptible to local minima stucking. The results of classification of signals with and without late potentials based on trained neural network models showed that training using scalograms obtained on the base of a modified wavelet transform allows achieving 99% classification accuracy, which is 1-49% more than that using scalograms obtained on the base of on the classical wavelet transform.Документ Відкритий доступ One-Bit Spectrum Sensing Peculiarities(КПІ ім. Ігоря Сікорського, 2024) Buhaiov, M. V.The use of autonomous sensors for analyzing wide bands of radio frequency spectrum requires using the highspeed analog-to-digital converters (ADC). The operation of such devices with high bit resolution requires several watts of power. One-bit ADC will reduce power consumption and simplify the analog part of the receiver. The goal of the research is to determine the features and conditions under which it is advisable to analyze the radio frequency spectrum in a sophisticated radio electronic environment using a one-bit ADC. The problem of using such ADC is the deterioration of the signal-to-noise ratio (SNR) and the occurrence of spurious spectral components at its’ medium and high values. The task of spectrum sensing is complicated by the fact that a large number of signals with different values of bandwidth and SNR can be simultaneously present in the analyzed frequency band. After a one-bit ADC, the energy of the complex signal remains unchanged regardless of its shape. Therefore, at high SNR values, the signal energy flows into spurious spectral components. When the bandwidth of the signal increases to 30% of analyzed frequency range and higher, the spurious spectral components are transformed into a broadband pedestal. Analytical dependencies have been obtained that allow us to calculate the values of the SNR at which signals can be detected and spurious spectral components appear. These dependencies take into account the parameters of the Welch periodogram and the level of spectrum occupancy. The results of the analysis of real signals showed that the shape of the signal spectrum after one-bit ADC repeats the shape of the spectrum after an 8-bit ADC, except for the absence of some weak signals. The cost for simplified hardware implementation and lower power consumption when using one-bit ADC compared to ADC with higher resolution is a narrowing of the dynamic range of signals. The use of one-bit ADC for spectrum sensing will be justified only when there is information on the analyzed bandwidth occupancy and the SNR of the signals that can be observed.Документ Відкритий доступ Perspective of Creating Low-Cost Medical Assistant Robot Based on Waffle PI4 Platform with Palm Vein Pattern Scanner(КПІ ім. Ігоря Сікорського, 2024) Anufriiev, V. V.; Levchenko, O. O.; Levchenko, Ye. V.; Chekubasheva, V. A.; Glukhov, O. V.; Galat, O. B.The purpose of this study is to develop a set of modifications for the TurtleBot 3 Waffle Pi robotic platform. One of the key achievements of this work is the creation of a biometric identification system based on the venous pattern of the palm. The principle of operation of the identification system is based on the use of infrared radiation absorbed by hemoglobin in the venous system of the palm. The absorbed radiation creates a clear pattern that can be captured using a camera without an infrared filter. The resulting image is pre-processed to reduce noise and unify with other images for further use in training a convolutional neural network used for patient identification. This identification method allows for high-speed and accurate patient identification, even with dirt or scratches on the palm. The described modifications are aimed at expanding the capabilities of the platform for military medical applications. By integrating these modifications into the TurtleBot 3 Waffle Pi robotic platform, military and civilian hospitals can improve their ability to provide timely and accurate medical care to those in need.Документ Відкритий доступ Scattering of Electromagnetic Waves by Frequency-Detuned Systems of Dielectric Resonators(КПІ ім. Ігоря Сікорського, 2024) Trubin, O. O.The general problem of scattering on a system of frequency-detuned coupled Dielectric Resonators (DRs) located in one or several transmission lines is considered. The field describing the natural oscillations of the system of detuned DRs is decomposed over the field of partial resonators. A system of equations is derived, the solution of which allows one to determine the frequencies and amplitudes of the system’s natural oscillations. It is shown that the resulting system of equations, by means of algebraic transformations, can be reduced to the problem of determining the eigenfunctions and eigenvalues of a finite-dimensional operator, determined through the elements of the coupling operator of frequency-detuned DRs. The limitations of the proposed calculation method are noted. The solution to the scattering problem is expanded in terms of the natural oscillation field of the system of detuned DRs. A system of linear equations for the amplitudes of forced oscillations is obtained. It is shown that in the particular case of identical resonator frequencies, the found system exactly coincides with the equations obtained previously for various types of DR. General solutions are found for the scattered field on frequency-detuned resonators located in different transmission lines. Several examples are given of calculating the frequency dependences of the scattering matrix for the most interesting structures consisting of coupled dielectric resonators detuned in natural oscillation frequencies. The frequency characteristics of the scattering of two bandstop filters with different stop bands, made on detuned DRs in the same transmission line, are calculated. The frequency dependences of the scattering matrix on two bandpass filters located parallel between regular lines are calculated. The scattering characteristics of bandpass filters of complex designs containing various DRs are calculated: a bandpass filter built on a system of coupled DRs in a transmission line break and several DRs in a regular line, as well as an elliptical bandpass filter. The capabilities of the proposed method are demonstrated by the example of optimizing the scattering characteristics of a demultiplexer built on the basis of two bandpass filters with different passband frequencies. The proposed method also makes it possible to calculate the scattering characteristics of bandpass and bandstop filters with several operating frequency bands used in modern communication systems.Документ Відкритий доступ Scattering of Electromagnetic Waves by Loss and Gain Systems of Dielectric Resonators(КПІ ім. Ігоря Сікорського, 2024) Trubin, A. A.The problem of wave scattering on a system of coupled Dielectric Resonators (DR) made of an active or absorbing dielectric is considered. The solution of the scattering problem is decomposed over the field of natural oscillations of the DR system. The field describing the natural oscillations of the DR system is decomposed by the field of partial resonators, which are made of a dielectric with a complex permittivity. A system of equations is given, the solution of which allows to determine the frequencies and amplitudes of the natural oscillations of the system of active or absorbing resonators. In the work, a new system of linear equations for amplitudes of forced oscillations of resonators was obtained. General solutions for the scattering field on resonators located in a regular transmission line or in a break of a regular line have been found. Several examples of calculation of the frequency dependences of the scattering matrix for different bands to pand band-pass filters, consisting of coupled active or absorbing dielectric resonators are given. The possibilities of the proposed method are demonstrated on the example of optimization of scattering characteristics on band-stop and band-pass filters made of an active dielectric. It is shown that the use of resonators made of an active dielectric will make it possible to build and optimize the frequency characteristics of a new class of devices that simultaneously perform the functions of filters and amplifiers. The conditions under which it is possible to build filters with the functions of amplifiers are defined. In the future, the proposed devices may find application in optical communication systems.Документ Відкритий доступ Аlgorithm for Spectrum Sensing and Signal Selection by External Parameters(КПІ ім. Ігоря Сікорського, 2024) Buhaiov, M. V.For modern radio monitoring, a panoramic view of a wide frequency band and signal selection is its most important part. The constant growth of the number of radio electronic devices and the expansion of the instantaneous bandwidth of analysis in modern radio receiving devices leads to the fact that a significant number of analog and digital signals can be observed at the same time. Automatic adaptation of radio monitoring system to further signal processing is possible due to preliminary signal selection. The goal of this research is to develop an algorithm for signals selection in panoramic radio monitoring systems by their external parameters. The essence of proposed algorithm is to detect occupied bands of radio frequency spectrum, estimate center frequency and bandwidth of each channel, noise level and signal-tonoise ratio. Creation of frequency channels allows for signal filtering and estimation of pulse durations, as well as occupancy of each channel. Estimates of parameter for each signal fragment and frequency channel are recorded in associative arrays, which simplifies further signal selection. Due to variability of noise and propagation channel, estimates of signal parameters for each signal fragment are random variables. To obtain reliable estimates of signal center frequency and bandwidth, they are further grouping. Array of data can be accessed both by frequency channel number (table rows) and by signal parameters (keys), which are table column headers. Associative relationships between data provide flexible signals filtering by any combination of parameters. To test developed algorithm, we analyzed frequency band of 933-953 MHz and used the DataFrame Multi Index data container of Pandas package of Python programming language. This structure provides multi-level indexing, flexible access to data, and a wide range of tools for their processing and modifying. Developed algorithm can be used in existing and future radio monitoring systems for radio electronic devices identification and databases creation.Документ Відкритий доступ Адаптивне виявлення сигналу рухомої цiлi в FMCW радарi при невiдомiй потужностi шуму(КПІ ім. Ігоря Сікорського, 2024) Маленчик, Т. В.; Жук, С. Я.Одним з перспективних засобiв виявлення рухомих цiлей на малих вiдстанях є FMCW радар. Вiн забезпечує: високоточне вимiрювання дальностi, а також радiальної швидкостi, низьке енергоспоживання. Вихiдною iнформацiєю для алгоритмiв виявлення сигналiв цiлей в FMCW радарi є далекомiрнодоплерiвський портрет. Вiн формується шляхом двомiрного дискретного перетворення Фур’є (ДПФ) над отриманими на iнтервалi когерентного накопичення демодульованими сигналами вiдповiдних перiодiв модуляцiї. У випадку однорiдного шуму з невiдомою потужнiстю викорстання алгоритмiв CFAR призводить до надлишкових обчислювальних витрат за рахунок ковзного оцiнювання потужностi шуму. Крiм того розмiри ковзного вiкна обмеженi, що не дозволяє отримати оцiнку потужностi шуму з необхiдною точнiстю. В якостi математичної моделi корисного сигналу вiд цiлi може бути використаний гармонiйний сигнал з невiдомими амплiтудою, частотою i початковою фазою. Розглянуто алгоритм адаптивного виявлення гармонiйного сигналу з невiдомими параметрами, отриманого на iнтервалi когерентного накопичення FMCW радара при вiдомiй потужностi шуму. Пристрiй виявлення побудовано за перiодограмною схемою. Проведено аналiз характеристик виявлення сигналу FMCW радара при вiдомiй потужностi шуму, якi можуть виступати в якостi нижньої границi при невiдомiй дисперсiї шуму. На основi методу максимальної правдоподiбностi запропоновано алгоритм оцiнювання невiдомої потужностi шуму на основi тестової вибiрки, отриманої з далекомiрно-доплерiвського портрету. Оцiнка невiдомої потужностi шуму є вибiрковим середнiм. На основi методу iнтервального оцiнювання визначено довiрчi iнтервали щодо ймовiрностей хибної тривоги i виявлення цiлi в залежностi вiд об’єму тестової вибiрки. Границi ймовiрностi хибної тривоги не залежать вiд оцiнки потужностi шуму. На основi отриманих залежностей можна визначити об’єм вибiрки, що забезпечує допустиме значення довжини довiрчого iнтервалу ймовiрностей хибної тривоги i виявлення цiлi.Документ Відкритий доступ Алгоритм пiдвищення точностi прогнозування вихiдної потужностi сонячної електростанцiї при застосуваннi методу найближчих сусiдiв(КПІ ім. Ігоря Сікорського, 2024) Седляров, Є. В.; Клен, К. С.У статтi наведено метод прогнозування вихiдної потужностi сонячної електростанцiї на основi методу регресiї найближчих сусiдiв. Для створення прогнозу вихiдної потужностi використано данi з сонячної електростанцiї в Китаї потужнiстю 50 МВт, що має данi значень сонячної радiацiї, навколишньої температури, тиску та вологостi i вiдповiднi їм значення вихiдної потужностi в перiод з 01.01.2019 до 31.12.2020 з дискретнiстю 15 хвилин. Перед використанням даних проведено їх очищення вiд викидiв методом стандартного мiжквартильного розмаху, роздiлення точок даних на тестову та тренувальну групи i застосовано масштабування ознак методом стандартизацiї для коректного обчислення Евклiдової дистанцiї мiж точками даних. Запропоновано застосування кластеризацiї погодних параметрiв методом k-середнiх, що дозволяє провести iндивiдуальний вибiр кiлькостi сусiдiв для кожного кластеру i виключити вплив точок одного кластеру на кiлькiсть сусiдiв iншого кластеру. Кiлькiсть кластерiв обирається за допомогою визначення коефiцiєнта силуету, тренувальна група погодних параметрiв з вiдповiдними їм значеннями вихiдної потужностi розподiляється по кластерам на основi Евклiдової вiдстанi до центроїд кластерiв. Тестова група погодних параметрiв розподiляється по кластерам, пiсля чого прогнозування методом найближчих сусiдiв вiдбувається в межах кожного кластеру окремо. Вихiдна потужнiсть сонячної електростанцiї розраховується як середня арифметична зважена потужностей сусiдiв кожної точки тестової групи. В кiнцi алгоритму послiдовнiсть точок погодних параметрiв тестової групи вiдновлюється i створюється часовий ряд прогнозу вихiдної потужностi. Запропонований алгоритм дозволив зменшити похибки прогнозу MSE, RMSE, MAPE, MAE на 1 день на 0,5348, 0,2265, 0,38%, 0,1448 вiдповiдно, на 7 днiв похибки стали меншими на 0,1992, 0,0384, 0,1%, 0,0193 вiдповiдно. В результатi вiдносна похибка при прогнозуваннi на 24 години складає 4,22%.Документ Відкритий доступ Алгоритм послiдовного виявлення траєкторiї малорозмiрного БпЛА FMCW радаром за критерiєм «найсильнiший сусiд»(КПІ ім. Ігоря Сікорського, 2024) Маленчик, Т. В.; Жук, С. Я.Для забезпечення прийнятної ймовірності виявлення малорозмірних БпЛА виникає необхідність зниження порога виявлення, що призводить до суттєвого збільшення ймовірності хибної тривоги в елементі розрізнення (більше 10-3). Для підвищення ефективності вирішення задач вторинної обробки радіолокаційної інформації при збільшенні кількості хибних відміток, використовуються вирішальні статистики відміток, отримані при первинній обробці сигналів. Відомі алгоритми послідовного виявлення траєкторії цілі з використанням вирішальних статистик відміток вимагають значних обчислювальних витрат. Для вирішення задачі виявлення траєкторії цілі використовується послідовний критерій відношення правдоподібності Вальда з постійними порогами, що розраховуються на основі заданих ймовірностей правильного і хибного виявлення траєкторії цілі. Отримано математичний вираз часткового відношення правдоподібності, у якому враховано щільності ймовірності вирішальної статистики відмітки за умови, що вона є цільовою або хибною, а також ймовірності: виявлення цілі і хибної тривоги в елементі розрізнення, попадання цільової відмітки в строб підтвердження траєкторії, відсутності в стробі хибних відміток. Аналіз запропонованого алгоритму та його порівняння з відомим, у якому ототожнення відміток відбувається за критерієм близькості до центру стробу, проведено за допомогою статистичного моделювання на прикладі виявлення траєкторії цілі за даними FMCW радара, що вимірює дальність та радіальну швидкість цілі. Для опису щільностей ймовірності вирішальної статистики відмітки за умови, що вона є цільовою або хибною використовуються нецентральний і центральний розподіли хі-квадрат з двома ступенями свободи. Для розглянутого прикладу, на відміну від відомого алгоритму, в якому не враховуються вирішальні статистики відмітки, розроблений алгоритм забезпечує збільшення ймовірністі виявлення траєкторії цілі при α = 10-2, 5 x 10-3 на 14%-50% і 4%-34% відповідно. При цьому середнє число оглядів при α = 10-2, 5 x 10-3 зменшується приблизно в 4 та 2.5 рази. Ймовірність хибного виявлення траєкторії цілі FΣт для розробленого алгоритму менше ніж на порядок. При цьому середнє число оглядів при α = 10-2, 5 x 10-3 зменшується приблизно в 3.8 та 2.3 рази відповідно.