Home – Home – Jornals – Avtometriya 2025 number 2

The use of microgrids is a relevant direction of energy development both in Russia and abroad. At the same time, problems arise in selecting an optimal structure and operating parameters of such networks. Conducting full-scale experiments with a real energy complex involves certain risks associated with high time and financial costs, as well as the reliability and safety of its equipment. Therefore, studies are usually carried out on mathematical or physical models of a microgrid. A physical model allows solving a number of important problems. These include the analysis of the network structure, the study of its main characteristics (resilience, reliability, regulation of demand and supply of energy resources, use of renewable energy sources, and other features) and the determination of optimal operating modes. The paper considers a monitoring system for a physical model of a microgrid (a simulation stand that simulates a network with a high share of renewable energy sources).

A method and technical instrumentation are proposed for estimating the sizes of vortex structures generated in the annular gap of a Couette - Taylor system. The size estimation of these structures is based on analyzing the measured amplitude-frequency spectra of radial dynamic pressure pulsations. A simplified model of vortex generation in the Couette - Taylor annular flow, driven by the rotation of the inner cylinder surface, is introduced. Using this model and the measured amplitude-frequency spectra of radial pressure pulsations, the sizes of the generated vortices in the Couette - Taylor flow are evaluated. The results reveal three distinct types of vortices: vortices with diameters significantly smaller than the annular gap width, vortices with diameters comparable to the gap width, and vortices with diameters substantially exceeding the gap width, leading to their deformation (flattening).

The paper presents a gradient boosting-based machine learning model developed to predict the occurrence of defects (dross) on galvanized steel sheets used in the automotive industry. An analysis of the influence of process parameters on the occurrence of defects is carried out, which makes it possible to identify the key factors affecting the quality of the coating: coil rolling speed, elongation and force in the coil-scin pass mill, zinc coating at top, temperature in the galvanizing pot, and temperature in the furnace snout. Based on the results of the study, a digital assistant is developed for evaluating coils in real time, providing simulation of the decision-making process and helping in prompt managing of the technological process.

The article presents modifications to the Sardana software package for automating synchrotron radiation experimental stations. These modifications are aimed at achieving the nominal accuracy of positioners, reducing the time consumption with the minimum operator supervision. In addition, these modifications simplify conducting standard experiments and provide greater flexibility in configuring experimental setups. They are applied to the automated control system of the sample environment at Experimental Station 1-1 "Microfocus."

The article problem of optimal performance for a process described by a system of nonlinear differential equations is considered. To determine an approximate solution to the problem, it is proposed to apply the evolutionary method of artificial immune systems, which offers an advantage of the lack of sensitivity to the choice of the initial approximation. To find a numerical solution to the problem of optimal performance, a step-by-step algorithm is formulated, whose operation is tested on a model problem. Suboptimal control in terms of response time is obtained, and the algorithm parameters are determined that allow one to calculate the problem solution with the least computational costs. The efficiency of the developed algorithm is demonstrated by comparing the results of numerical calculations with the results of applying the fixed-point method.

A parametric identification problem of a linear dynamic SISO plant under incomplete a priori information based on the least squares method is considered At the stage of formation of the system of equations, a special type of the forming function is used. It allows the numerical differentiation of the measured input and output signals of the identified plant to be replaced by multiple application of integration by parts. The dependences of the frequency properties of the forming function and its derivatives on the function parameters are obtained in analytical form. The paper presents results illustrating almost complete coincidence of the frequencies obtained both experimentally and analytically.

We develop a theory of the ferroelectric polarization reversal taking into account conduction of domain walls. The case of capacitor geometry is considered as applied to lithium niobate. We show that the field dependence of the domain nucleation rate obeys the law exp(-E_n/E ), where E is the applied field and E_n ≈ 10² kV/mm is the characteristic field controlling the domain nucleation process. For the critical domains, the longitudinal size l^*_c strongly exceeds the transversal size 2l^*_a ≈ 1 nm. We formulate consistently a kinetic model of the field-induced polarization reversal. It includes not only random nucleation events but also the subsequent elementary events of the lateral growth obeying the Merz law exp(-E_l/E ) with the field E_l = E_n/3√3 controlled by the crystal symmetry. Numerical simulations show distinct stages of the domain nucleation, lateral growth, and coalescence. In accordance with experiment, the polarization switching time obeys the law exp( E_*/E ) with E_* = (E_n + 2E_l)/3, while the hysteresis loops show a standard behavior with the coercive field E_c = (3-5) kV/mm weakly dependent on the field ramping period.

Fluorescence spectroscopy, which employs excitation wavelength scanning, is a powerful tool for chemical, biological, and medical diagnostics. It allows one to characterize the distribution of fluorophores in samples of various nature. However, when the fluorophores are initially unknown, interpreting large sets of excitation wavelength-dependent fluorescence spectra can be challenging, especially when there are data losses in the measured excitation-emission matrices due to random causes or systematic factors. In this paper, we propose to restore the damaged areas of the spectrum using the principal component analysis. This can be achieved without data loss because individual fluorophores’ fluorescence spectra are independent of the excitation radiation wavelength.

The present paper describes the Monte Carlo simulation of the silicon nanowire growth by the vapor-liquid-solid mechanism using gold droplets as a catalyst. The growth of tilted nanowires on Si surfaces with (111), (011) and (100) orientations is considered. It is found that the silicon nanowire orientation and morphology is affected by the size of gold droplets. The growth of tilted nanowires in the ⟨011⟩ and ⟨211⟩ directions using small-size droplets is demonstrated. As the droplet size increases, ⟨111⟩ becomes the most probable direction of nanowire growth. Diagrams demonstrating the influence of the gold droplet size on the probability of the nanowire growth direction on Si surfaces are presented. The kinetics of unstable growth of silicon nanowires catalyzed by small-size droplets is analyzed.

A method has been developed for modeling the transmission of channel signals of the M-QAM modulation format using the CO-OFDM (Coherent Orthogonal Frequency Division Multiplexing) technology over a fiber-optic path operating in a nonlinear regime. Using the normalized analytical expressions obtained, signal distortions on the receiving side are modeled for given parameters of the transmission system, and the signal quality factor (Q-factor) is calculated based on a statistical analysis of the signal and noise parameters. The model helps to clearly represent the signal distortion in the constellation of signal points in the IQ-diagram for given parameters of the transmission system. The paper presents a model of a transmission system for an OFDM signal with a total rate of 4 Tbit/s using a standard single-mode optical fiber over distances of 80, 160, and 240 km. To achieve the maximum signal quality factor at such transmission distances, optimal levels of optical signal power of +5.5, +4, and +2.5 dBm have been found.

Image quality assessment is crucial in the design of optical systems with an increased depth of the field and containing axicons, which possess two caustics, retain the linearity property but are spatially variant, creating difficulties for their analysis. We propose to consider the images formed by each of the caustics within the framework of the linear systems theory. A theoretical analysis of image formation is performed and PSFs are found for three optical systems in which toroidal light waves are formed by means of the axicon, as well as multiple experiments are carried out to determine the resolving power of such systems. As a result, it is shown that the axicon produces an image of a single object as a continuous set of ‘intermediate’ images along the focal segment. Light beams incident from the object to the centre of the axicon produce an image in the immediate vicinity of the axicon, while beams incident to the periphery of the axicon produce an image at the periphery of the focal segment. In doing so, all these low-contrast images mask each other. The lens-axicon tandem simultaneously produces images of the object both at the near diffraction zone in the form of a set of ‘intermediate’ images and at the far zone in a geometrically transformed form. Moreover, in the near diffraction zone, the depth of the field imaged by the tandem can exceed the depth of the field imaged by the single objective. The experimental data confirm the theoretical conclusions.

The paper presents the study of the transmission loss dependence in the 1.5-1.6 μm range for tapers, depending on their length, melting speed, and constriction dimensions using the Corning SMF-28 optical fiber as an example. The tapers are fabricated on the Fujikura FSM-100P, 100P+, LZM-100 welding machines and Lightel CW-200B coupler workstation. As a result of the experiments, a taper configuration with the minimum losses of -0.5 dB is determined. Using the Lightel CW-200B station, it is possible to manufacture a mode adapter for matching the modes of a standard polarization-maintaining optical fiber Fujikura SM15-PS-U25D and a highly nonlinear fiber with a reduced core HNLF PM. Due to better matching of the mode fields, the connection losses of these optical fibers are reduced from -3 to -1.25 dB.

The approach of applying machine learning methods to automatically detect people in forest-steppe areas in images is investigated. The artificial neural network technology is used to detect people from photographs. Deep learning methods of convolutional neural networks (YOLOv51, DenseNet, CenterNet, InceptionV3, Xception, and Faster R-CNN) are used in interaction with the "transfer learning" technique. Based on YOLOv51, a neural network is trained, which makes it possible to identify people using graphical images with an accuracy of 0.8795 on a test sample using the F1-score metric with a threshold value of 0.5.