ISSN 2224087X (Print) ISSN 22240888 (Online)
Collected scientific papers "Electronics and information technologies"
(In 19662010 published under the title "Electrical engineering")
(Certificate of State Registration ÊÂ ¹176186468ÏÐ from February 11, 2011)
Issue 12, Pages: 319 
DOI: https://doi.org/10.30970/eli.12.1 
DETECTION OF TECHNICAL FAILURES ON PRODUCTION LINES USING MACHINE LEARNING, LINEAR AND BAYESIAN MODELS OF LOGISTIC REGRESSION 
Bohdan Ì. Pavlyshenko 
In this work, we study the use of logistic regression in manufacturing failures detection. As a data set for the analysis, we used the data from Kaggle competition "Bosch Production Line Performance". We considered the use of machine learning, linear and Bayesian models. For machine learning approach, we analyzed XGBoost tree based classifier to obtain high scored classification. Using the generalized linear model for logistic regression makes it possible to analyze the influence of the factors under study. The Bayesian approach for logistic regression gives the statistical distribution for the parameters of the model. It can be useful in the probabilistic analysis, e.g. risk assessment. 

Issue 12, Pages: 2029 
DOI: https://doi.org/10.30970/eli.12.2 
GENERATION OF ELEMENTARY SIGNS IN THE GENERAL SCHEME OF THE RECOGNITION SYSTEM BASED ON THE LOGICAL TREE 
I. Povhan 
In this work consider an important problem of simple and effective generation of elementary features in the general scheme of the approach  branched feature selection in the construction of treelike schemes of recognition of discrete objects. Since the correct choice of features in the process of constructing the recognition system affects the simplicity and efficiency of the resulting scheme, a simple method of generating elementary features is proposed and several simple algorithmic implementations effective in terms of memory saving and performance are given on the basis of it. 

Issue 12, Pages: 3038 
DOI: https://doi.org/10.30970/eli.12.3 
FACIAL RECOGNITION WITH USING OF THE MICROSOFT FACE API SERVICE 
Volodymyr Grabovskyi, Oleh Martynovych 
Results of creation and testing of software for recognition and comparison of human faces are presented. The program was created to run on Windows and implemented using the features of the face recognition service Microsoft Face API. Some peculiarities of work and practical use of this service for creation of user programs are analyzed. The work of the created program was tested on real examples of face recognition both on images belonging to one person and to different persons. The influence of the presence of emotions on the face and different shooting conditions of the submitted photo on the recognition result is investigated. Some aspects of the practical use of this service are also noted. In particular, the positive aspect is the ability to allow the use of Face API services for free for 30 days, which allows users to develop and test a recognition program and evaluate the quality of its functionality before acquiring commercial rights to the Microsoft API. 

Issue 12, Pages: 3948 
DOI: https://doi.org/10.30970/eli.12.4 
LONG ARITHMETIC IN EXCEL. III. SOLUTION OF SYSTEMS OF LINEAR ALGEBRAIC EQUATIONS FOR TEST PROBLEMS OF THE FINITE ELEMENT MODELLING 
V. Fourman, M. Khomyak, Ya. Marko 
The class of rational numbers and full precision algorithms are adopted to solve systems of the linear algebraic equations. The simple method of Gaussian elimination is implemented which shows typical computational procedures and costs of computer resources, especially the time of calculus. One practical application of this algorithm is test tasks solving and software verifying. For example, the finite element method is often used in the solid mechanics modelling. Brief theoretical basis of the variational problem of bending of a complaint to shear plate under constant pressure is given. The perspective scheme of the FEM using Bsplines is proposed. For example, onedimensional finite elements using quadratic opened Bsplines as basis functions are considered. Typical matrices contain the product of the basis functions and their first derivatives in various combinations under the integral. In the case of a single length of these elements, the matrices are obtained analytically in the form of rational numbers. The assembly procedure for ten elements with two degrees of freedom on each is illustrated. The sparse global matrix and right handle size of the resulting system with rational coefficients are constructed and its exact solution is obtained. Also examples of filled symmetric matrices are considered. A case of rigid systems with illconditioned Gilbert’s matrices dimensionality up to N=200 is examined. The maximum number of digits of a denominator is 240 for N=200, what explains the inaccessibility of exact solutions when using double precision. Numeric simulation confirmed the expected analytical precision and the robustness of the proposed program code. In particular, the time cost and increasing of rational coefficients size during the direct way of the Gauss method are traced. We compared a "slow" and "fast" computers with 2 processors and 2 MB of RAM vs. 4 processors and 8 MB of RAM respectively. The calculation time was reduced by three times. This indicates that significant computer resources are critically required. On the other hand, long arithmetic using binary algorithms can be implemented more efficiently. 

Issue 12, Pages: 4954 
DOI: https://doi.org/10.30970/eli.12.5 
WEBAPPLICATION FOR MANAGING YOUR OWN FINANCES AND CURRENCY CONVERSION 
O. Babych, V. Venhrov 
Moneta is an application, that helps people to manage their finances. Web applications have been developed using the Django framework for building websites. The software implementation allows you to create, edit and delete both wallets and expenses, as well as the ability to share your own wallets with other users. It is also possible to view transaction statistics for any period of time. It uses modern web development tools, such as Django Framework, jQuery. The development environment was chosen PyCharm Community Edition, and the server side was written in Python. As a database management system, MySQL was used. When registering, the user avoids entering a minimum of data: email, default currency and password. When entering the data, the user is checked for the specified email in the database and for the entered passwords. After successful registration, the user can safely log in, but such a implemented functionality for password recovery, which is nowadays very relevant. An email is entered for checking the database, if the specified email exists, the user sees a message asking him to check his email. After that, the user password in the database is changed to a random set of 10 characters, and a letter with a new password is sent to the user by email. The user can view their transaction statistics for any desired period of time. By default, statistics for the current month and the current year are displayed. Revenue statistics are presented in different graphs to avoid confusion. Regardless of the currencies in which the user’s transactions were made (receipts / expenses), when statistics are displayed, all transactions are converted to the default currency for the user, and the statistics will be displayed in that currency. The exchange rate is taken from the National Bank’s API. The user can change their login details, such as password and default currency, an unlimited number of times. If the profile is deleted, the user will not be able to recover it, all data will be lost, however, if the user shared his or her wallet or expense, the owner will be transferred to that user after deleting the profile. 

Issue 12, Pages: 5563 
DOI: https://doi.org/10.30970/eli.12.6 
METHOD OF SPACE IMAGE IMPROVEMENT BY USING SPATIAL OPTICAL MASK AND FREQUENCY FILERS 
I. Polovynko, A. Kashuba 
Simple but effective program for the improvement of quality of space images, which is based on spatial and frequency filtration with a variety of filters, is provided are present. The theory of creating optical masks, which consists in using a convolution operation, is described. As an example, Gauss smoothing filters are considered. Practical part of this work was realized by using Python 3.7.4 program language in the environment Visual Studio 2017 Community Edition. Some programs where realized by using Jupyter Nonebook. This are presented 8 spatial filters. Also frequency filtering is performed by using Lowpass and Highpass filters at different fitting parameters. As an example for implementation of this program, the image space photography of Defense monastery of St. Basil the Great was taken. The results of filtering the image in the frequency range by using Lowpass and Highpass filters at different fitting parameters are shows, that it is better to apply frequency filtering after spatial selection in certain areas. In particular, for the case of image filtering with Highpass filters, remains of lost defensive walls of Defense monastery of St. Basil the Great are clearly visible. 

Issue 12, Pages: 6471 
DOI: https://doi.org/10.30970/eli.12.7 
FORECASTING PROPERTIES OF CARBON MATERIALS WITH THE USE OF NEURAL NETWORKS 
R. Lisovskyi, I. Poplavskyi, B. Rachii, Z. Liubun 
Nanoporous carbon materials (NCM) due to the large specific surface area and unique physical and chemical properties are widely used in various branches of science and production as catalytic, electrochemical and sorption materials. Taking into account the widespread application and despite the deep and long history of their study carbon materials are very interesting research object, because of the quite large variety of raw materials, activation methods and chemical treatment of the surface. That’s why defining of optimal parameters and conditions of technological processes of chemical activation and temperature treatment of plant raw materials for porous carbon materials obtaining with predetermined parameters of the porous structure is an urgent scientific and practical task. Optimization of the methods for nanoporous carbon materials obtaining requires a large number of experiments that require significant material and time costs. Therefore, the development of a mathematical model for the dependence of the characteristic of NCM porous structure on the technological conditions of obtaining is an urgent task. The difficulty of physical processes in the investigated material causes significant difficulties in implementing a mathematical model with the help of mathematical dependencies in the form of a system of equations. Using the neural networks gives the possibility to avoid this problem. The neural network structure is taken and it is considered that after training it will play the role of a mathematical model of a physical object which then can be used for prediction. The created neural network plays the role of an approximator, which makes it possible to determine the properties of the physical system at arbitrary values. The ability of neural networks to approximate multidimensional functions is based on Kolmogorov’s theory. According to Kolmogorov’s theorem the only one hidden with the number of neurons 2N+1, were N dimension of function is sufficient for approximation of an arbitrary function. Numerical experiments have shown that a multilayered neural network can be used to predict the physical properties of nanoporous carbon materials. 

Issue 12, Pages: 7281 
DOI: https://doi.org/10.30970/eli.12.8 
THE BEHAVIOR OF THE LYAPUNOV’S EXPONENTS FOR THE INCOMMENSURATE STRUCTURE WITH THE PARAMETER OF THE SPONTANEOUS DEFORMATION AS THE ORDER PARAMETER 
S. Sveleba, I. Katerynchuk, I. Kunyo, I. Karpa, Ya. Shmygelsky 
The Lyapunov’s exponents in a wide range of parameters that determine the dynamics of an incommensurate superstructure for ferroelectric crystals $mathrm{[N(CH_3)_4]_2CuCl_4}$, which have an incommensurate superstructure were calculated. The calculation of the Lyapunov’s exponents was performed by the AdamsMulton method in the Python software environment using the JiTCODE library. It is established that the positive value of one Lyapunov exponent, and the negative value of the other three exponents for incommensurate superstructure, is characterized. The incommensurable superstructure is characterized by a strange attractor with a boundary cycle, since the third indicator acquires a value that far exceeds the sum of all others. The Lyapunov’s exponent spectrum, with a constant positive value of the first Lyapunov’s exponent for an incommensurate superstructure described by a twocomponent order parameter, is characterized. The strongly degenerate abnormal behavior of Lyapunov’s third and fourth exponents shows that the incommensurable superstructure is characterized by hyperchaos, which, according to the authors, describes the appearance of a chaotic phase, and the establishment of a quasistable state with the appearance of longperiodic proportional phases. The ground state of the system is characterized by a boundary cycle, as evidenced by the Fourier spectra of an incommensurate superstructure. Fourier spectrum of the boundary cycle is discrete with distinct peaks at frequencies that correspond to the fundamental harmonics of the cycle. In contrast to the boundary cycle, the distribution of the spectral density of the chaotic attractor, which occurs at T = 1.0 and K = 2.0, is continuous, but it preserves all peaks, which, conventionally speaking, are a "memory" of the harmonics of the missing boundary cycle. They are clearly distinguished in a continuous Fourier spectrum. Chaotic attractor is characterized by anomalous spatial behavior of the amplitude and phase of the order parameter. The study of phasefrequency characteristics of spatial oscillations of the amplitude of the order parameter were carried out. Gradual disappearance of the periodic structure as the magnitude K (0 < K < 1.3) increases, the phasefrequency characteristic becomes chaotic in the condition of transition to chaos (at K $geq$ 1.5), and its autocorrelation is a kind of "white noise". 

Issue 12, Pages: 8291 
DOI: https://doi.org/10.30970/eli.12.9 
THE FEATURES OF BEHAVIOR OF THE LYAPUNOV’S EXPONENTS FROM THE SYMMETRY OF THERMODYNAMIC POTENTIAL DESCRIBED BY LIFSHITZ INVARIANT 
S. Sveleba, I. Katerynchuk, I. Kunyo, I. Karpa, Ya. Shmygelsky 
The influence of the thermodynamic potential (n) symmetry on the behavior of the Lyapunov exponents is investigated. The calculation of the system of secondorder differential equations was performed by a multistep variableorder BDF method. The Lyapunov coefficients were calculated using Python software using Skipy, JiTCODE libraries. It is established that for values of n $geq$ 5, the dynamics of the incommensurate superstructure is determined by the spatial change in the amplitude of the order parameter (the first integral of the motion of the incommensurate superstructure). For n < 5, the dynamics of a incommensurate superstructure is determined by the spatial change of the phase of the order parameter and its perturbation. It is shown that the dynamics of the superstructure is not affected by the ferroelectric or ferroelastic nature of the incommensurate superstructure. In order to obtain a detailed picture of the dynamics of the incommensurate superstructure, phase portraits were investigated in the coordinates R, R’, and $varphi$ at different values of the anisotropic interaction of the superstructure and the initial phase values of the order parameter $varphi_0$. It is established that the dynamics of this system is determined by the cascades of transition to chaos and areas of selforganization of the superstructure. It is shown that at the final stage of the transition to chaos, the system is characterized by the emergence of a chaotic attractor, which has a more complex phase portrait structure. The projection of such an attractor consists of two symmetrical parts with respect to the axis of the abscissa. The movement of a typical trajectory of a chaotic attractor can be divided into two phases. In the first phase of the trajectory, chaotic movements are made along the turns of the upper (lower) part of the attractor from time to time, passing to the boundary of the localization region where the stability of the upper (lower) symmetric boundary cycles is lost. At an unpredictable time, the trajectory moves from the upper (lower) part of the attractor to its upper (lower) part. This process is repeated infinitely many times. Thus, the transition to chaos has features characteristic of the Feigenbaum scenario (infinite cascade of bifurcations of doubling cycles), and for mixing (unpredictable mixing between the upper and lower parts of the emerging chaotic attractor). The selforganization process can be associated with the localization of a wave vector of incommensurate superstructure at a commensurate higher order value. It is shown that the distortion of the order parameter phase at the edge of the crystal can be related to boundary conditions in the first approximation. According to the authors, such cascading transitions to chaos are related to the modes of existence of a incommensurate superstructure. The emergence of chaos in the third stage is caused by the transition of a incommensurate superstructure to its stochastic mode of existence. In a stochastic mode of existence of a incommensurate superstructure, a chaotic, possibly incommensurate phase arises, which is a prototype of this chaotic state. It should be noted that the initial conditions (boundary conditions) stabilize the system, thereby leading to the removal of the degeneracy of the system, and the transition of a incommensurate superstructure to a state characterized by the existence of longperiod proportional phases. 

Issue 12, Pages: 92101 
DOI: https://doi.org/10.30970/eli.12.10 
ELECTRICAL IMPEDANCE ANALYZER OF SENSOR CONVERTERS BASED ON POROUS SILICON 
P. Parandiy, R. Yaremyk 
A portable measuring and computing system for information signals processing of a multielement sensor matrix based on porous silicon the IC5933A (Analog Devices) specialized impedance spectral converter integrated circuit was realized. This system of electrochemical impedance spectroscopy has been tested for character change of individual profiles problems of sensory elements impedance characteristics in the process of adsorption of different types watercontaining gases molecules. During the investigation it was fixed that the selected scheme including of sensor elements and the nominal value of the resistance element sensor units Rsence = 150 KOm, the optimal frequency range of impedance measurement is in the range of 100 Hz  10 kHz. Experimental data confirm the existence of functional dependence impedance spectra and partial sensitivities of individual sensors to individual gas component. These dependences are essentially analog converter characteristic gas sensors. The analysis of the nature of the change of the complex transfer coefficient module of each sensor allows to identify and numerically describe the adsorption processes of different types of gases, since they are characterized by individual frequency responses. On the set of the obtained experimental data, a method of constructing a mathematical model of the transformative characteristics of each sensory element is proposed. The method is realized on the realtime registration of the sensor impedance responses the to the external influence and an iterative algorithm of polynomial approximation of the obtained data array. Software for 3D visualization of measurement results is developed. Accuracy is done by comparing the realelement reference signal and its counterpart in the analytical method. At the end of each approximation cycle, a constant of the maximum approximation error for the given array is formed. The proposed approaches and the results obtained demonstrate the fundamental ability to identify the types of gases and calculate their concentrations, and can be the basis for the construction of systems for the express detection of molecules of CO, $mathrm{CO_2}$, $mathrm{H_2S}$, smoke, alcohols, propane and other hydrocarbons in the studied gas environments, as well as for environmental and biomedical monitoring. 

Issue 12, Pages: 102114 
DOI: https://doi.org/10.30970/eli.12.11 
DEVELOPMENT OF MULTISPECTRAL RECORDING MEDIA FOR MULTILAYER PHOTOLUMINESCENT INFORMATION RECORDING 
Pavlo Anikin, Ievgen Beliak 
Urgency of the research. Optical information recording by forming of microrelief information elements on the substrate is considered as the most efficient method of longterm data storage. Target setting. However, density of the optical information recording does not meet the requirements of the modern digital recording systems because diffraction limit significantly reduce the resolution of optical systems. Actual scientific researches and issues analysis. Scientific research in the area of optical recording shows the high potential of multilayer photoluminescent media. Volumetric recording is much preferable than development of subdiffraction optical systems due to technical simplicity.Uninvestigated parts of general matters defining. Typical problems of multilayer photoluminescent recording are low signalnoise ratio and low readout signal level, which makes this method to be inappropriate one for longterm data storage.The research objective. In this paper it was proposed method of volumetric optical information recording in multilayer, optically homogeneous media with photoluminescent information elements which spectra of photoluminescence differs from layer to layer .The statement of basic materials. To determine the optimal parameters of multilayer photoluminescent storage with multispectral recording medium the mathematical model of the photoluminescent multilayer recording process was developed.Conclusions. The task of finding the optimal parameters of the multilayer photoluminescent storage with multispectral recording medium was solved by finding the maximums of the objective functions. 

Issue 12, Pages: 115127 
DOI: https://doi.org/10.30970/eli.12.12 
MICROPROCESSOR DEVICE DETERMINING DIRECTION OF SOUND TO IMPROVE TECHNICAL EQUIPMENT AT WEAPONRY 
V. Mitkov, H. Klym, I. Vasylchyshyn, I. Karbovnyk, I. Trach, R. Dunets 
The problem of designing microprocessor devices for determining the direction to the source of sound with high accuracy is considered. The basic developments of leading foreign and domestic researchers on similar solutions are analyzed. The microprocessor device has been designed to determine the direction to the sound source using a modern element base. The structural and functional schematic diagram of the proposed device is described in detail. The reliability of the scheme and the average time of device to failure are calculated. Several variants of the algorithms for calculating directions to the sound source are considered. The microphone amplifier circuit was simulated to test the device’s performance. 

Issue 12, Pages: 128132 
DOI: https://doi.org/10.30970/eli.12.13 
MAGNETIC CIRCUIT FOR MMWAVEBAND MAGNETRON WITH AXIAL CATHODE SUPPORT 
V. V. Zavertanniy, K. Ilyenko, T. Yatsenko 
We report composite permanent magnet and iron polepiece magnetic circuit for 8 mm waveband (35 GHz) nonincandescent cathode magnetrons operating in the spatialharmonic regime. This reduces required values of magnetic induction to affordable figures and allows one to design magnetron permanent magnet magnetic circuits that have simple system assembly, reasonable massdimensional characteristics, and low cost impact. We chose to assemble the outer rim (magnetized normally towards to and away from the zaxis in the left and right poles, respectively) of each pole from a number of identical segment blocks with the same external diameter of magnetic material. Total axial extent of the both pole assembly and its external diameter are varied to achieve optimal massdimensional characteristics; central inserts house the iron polepieces, which also serve as magnetic field concentrators in the work gap of magnetron. Standard distance between the left and right pole assemblies of the permanent magnet focusing systems for mm waveband magnetrons manufactured at the Vacuum Laboratory of Department of Vacuum Electronics of IRE NAS of Ukraine is 25 mm. As the engineering criterion, we adopted the principle of the equal specific contribution into the zcomponent of the magnetic induction at the center of the work gap from a given portion of magnetic material. Practically, for the given magnetic induction value at the center of the work gap (which was set at 0.85 T), we calculated the total volume of magnetic material (of the both magnet poles) for possible dimensions of the magnetic circuit and chose those with minimal possible total mass. Thus, we worked out the design of permanentmagnet magnetic circuit for millimeter wavelength range magnetrons, which ensures optimum massdimensional characteristics at a given value of magnetic induction necessary for their operation. This work is supported in part by Emerging Security Challenges Division in the framework of NATO Science for Peace and Security Programme (Grant G5195). 

Issue 12, Pages: 133145 
DOI: https://doi.org/10.30970/eli.12.14 
A POSSIBLE PHYSICAL MECHANISM FOR THE FORMATION OF THE EXPONENTIAL FUNDAMENTAL ABSORPTION EDGE IN INDIRECT GaP SEMICONDUCTORS 
V. Belyukh, B. Pavlyk, H. Danylyuk 
Based on a mutually complementary analysis of the results of optical studies of layered semiconductors $mathrm{Tl_2S}$ and GaSe and calculations of the energy band structure of these crystals, a physical mechanism is proposed for the formation of the exponential fundamental absorption edge (FAE) in indirect gap semiconductors. The essence of this mechanism is the sequential activation of indirect transitions of electrons from the valence band to the conduction band with increasing energy of quanta of electromagnetic radiation incident on the sample. Thus, the absorption coefficient $alpha$ is defined as the sum of the absorption coefficients due to the corresponding indirect transitions of electrons with absorption and emission of phonons. The number of such indirect transitions of electrons depends on the energy band structure of a particular semiconductor. The resulting superposition of indirect transitions leads to the formation of an exponential FAE. Such a model allowed us to explain the exponential shape of the FAE of $mathrm{Tl_2S}$ and GaSe crystals in the entire spectral range of measurements. The simulation results showed a good agreement between the theoretical optical absorption spectra and the experimental spectra of these semiconductors. The numerical values of the band parameters determined during the simulation are in satisfactory agreement with the values of these parameters obtained from calculations of the energy band structures of $mathrm{Tl_2S}$ and GaSe crystals. This gives grounds to use the proposed model as a peculiar method for determining the band gap, $E_{gi}$, in indirect gap semiconductors in the case of an exponential shape of the FAE. Based on the results of studies of the optical properties of a layered GaSe semiconductor, an assumption was made about the possible universality of such a physical mechanism for the formation of the FAE in indirect gap semiconductors. However, to verify this assumption, further studies of the optical properties of other semiconductors are necessary. 

© Ivan Franko National University of Lviv, 2011
Developed and supported  Laboratory of high performance computing systems
