Software & Systems

Articles of journal № 3 at 2022 year.

Order result by: Public date | Title | Authors |

11. A filtration model and parallel computing of the blood flow characteristics in the circulatory system in case of an occluding circular clot [№3 за 2022 год]
Authors: Konyukhov V.M., Khramchenkov M.G., Konyukhov I.V.
Visitors: 2287
The authors have mathematically described the blood flow in the circulatory system based on the gen-eral filtration equations at the presence of an occluding circular porous clot located on the vessel wall, taking into account clot destruction under the action of hydrodynamic forces. The change in the inter-nal forces occurring in the clot material leads to forming an inhomogeneous permeability distribution, which is a function of spatial coordinates and time. Under the assumption of a symmetric toroidal shape of the thrombus, the paper considers a flat two-dimensional problem in the cross section of the vessel and a clot. The authors have developed finite-difference schemes and iterative algorithms using parallel com-puting technologies to solve corresponding two-dimensional problem in a blood vessel and a blood clot with heterogeneous permeability. They have also carried out parallelization at the visualization of the hydrodynamic characteristics of the flows. The software package with the implemented numerical and algorithmical models enables computational experiments with simultaneous visualization of their re-sults. There is the analysis of the influence of the shape and the structure of a clot formed at the blood vessel wall on the main hydrodynamic characteristics determining the probability of a clot breakage from the wall. It is shown that the most dangerous are the blood clots which boundary forms an acute angle with the vessel wall. This is caused by to two factors: the occurrence of the rotary moment of the forces tending to the break of the blood clot front edge down the vessel wall, as well as by the action of a local pressure gradient at the clot base, which significantly exceeds the pressure gradient in the vessel outside the blood clot. At the same time, the clot permeability affects the total pressure drop and the maximal pressure gradient magnitude, but it practically does not change the direction of the fluid flow in the blood vessel.

12. Applying artificial neural networks in automatic control systems for magnetic levitation [№3 за 2022 год]
Author: Korobeynikov, A.G.
Visitors: 2192
This paper considers the problem of neural network identification in the automatic control system for magnetic levitation (neural network controller). This problem is due to the fact that at present, it is typ-ical to use a mathematical apparatus based on artificial neural networks when designing automatic con-trol systems for complex technical dynamic objects. Such networks have the following unique ad-vantages: the possibility of parallel computing; finding previously unknown relationships between in-put and output sequences of digital signals; more efficient control of nonlinear systems using nonlinear activation functions. In addition, the use of artificial neural networks sometimes removes the difficul-ties that arise when describing some problems in the form of analytical mathematical models. This paper describes the solution for the problem based on artificial intelligence methods – well-known recurrent artificial neural networks NARX and LSTM. These networks were trained using the backpropagation algorithm and the Levenberg–Marquardt method, which have good convergence. For this problem, the application of trained artificial neural networks to test data has shown some ad-vantage of the NARX network compared to LSTM. Moreover, the RSME rms error for the NARX net-work with 50 hidden layers for this problem is less than for the network with 100 hidden layers. The MATLAB R2021b system was a tool for setting the architecture of neural networks, their con-struction, training and testing. The obtained results are applied in designing automatic control systems for levitating objects. Currently, most of these systems are being developed and used to transport goods for various purposes using the effect of magnetic levitation.

13. Remote control of a robotic device using virtual reality [№3 за 2022 год]
Authors: Kugurakova V.V., Khafizov M.R. , Kadyrov S.A. , Zykov E.Yu.
Visitors: 2986
The subject of this paper is the possibility of remote control of a robotic device using the virtual reality technology. This concept is being developed in order to create a more native way to control a device remotely, as well as a more convenient image transmission for an operator, which is especially relevant in extreme situations. The paper presents the analysis of intelligent control system technologies and the examples of im-plementing virtual reality technologies in tele-robotics. This helped to highlight the overall structure of the remote control system and to show the main implementation problems of this project. This analysis has become a base for development and implementation of a universal hardware and software telecon-trol platform on the Unity platform. The paper presents the application architecture and considers the requirements initially imposed on it in order to bypass the wireless channel limitations in the task of the real-time robot control using VR headsets. There is an implemented control system for selecting the direction of moving a robotic unit using one of the VR controllers. The paper considers implemented algorithms for obtaining high-quality stereoscopic vision in virtual reality devices, as well as the technique for synchronizing the re-mote rotation of a stereo pair. Moreover, there is a description of a command transmission system. The experiments of the implemented system were conducted in order to verify if it meets the neces-sary requirements to obtain the proper quality. The authors have measured the response time of the system and minimized the delay in the communication channel. The developed software for robotic unit telecontrol using a virtual reality headset provides the pos-sibility of stable telecontrol of remote robotic equipment.

14. Developing software for mathematical modeling of temperature distribution in the process of electron beam welding [№3 за 2022 год]
Authors: S.O. Kurashkin , Rogova D.V., Tynchenko V.S., Shutkina E.V.
Visitors: 2071
The purpose of the study is to develop a software system for modeling temperature distribution in the electron-beam welding process (EBW) in the Embarcadero RAD Studio application (student version), in the programming language C++. The base for this development is the thermal field theory using mathematical models to calculate the surface temperature distribution. Nowadays, it is possible to carry out simulation in the numerical simulation environments Comsol and ANSYS, as well as in the MATLAB package. These programs are powerful solutions, however they have a number of significant drawbacks such as: the need to create a product sketch in CAD systems and programming calculation models in the Comsol and ANSYS packages; direct work with the source code in the modeling process and lengthy calculations in the MATLAB package. Thus, the authors propose developing a software system based on mathematical models that will al-low: (1) simulating the EBW welding process for products from different alloys considering their geo-metric dimensions and thermophysical parameters, which will make it possible to determine the opti-mal technological parameters and, therefore, eliminate root defects; (2) simplifying the process of ELW modeling comparing to such systems as MATLAB, ANSYS and Comsol. The paper describes a software system for modeling temperature distribution, an algorithm for mod-eling an experiment, and an algorithm for calculating temperature. In addition, there is a presented modeling process using data based on full-scale experiments (technological parameters, thermophysi-cal parameters and geometric dimensions of the product). In addition, the developed software system allows storing both the results of the developed technological process and the simulation results. Introduction to production of the proposed approach to modeling the process of electron beam welding for thin-walled structures will reduce material and labor costs when developing the technolog-ical process of electron beam welding, as well as when introducing new types of products into produc-tion.

15. Applying CUDA technology for training the Kohonen neural network [№3 за 2022 год]
Authors: Latypova D.S., Tumakov D.N.
Visitors: 3044
The paper presents clustering of the data from in the training samples of the MNIST and Fashion MNIST databases. For clustering, the authors use a Kohonen neural network with a Euclidean metric for estimating distances. The optimal number of clusters (no more than 50) is determined for each handwritten digit (MNIST) and type of clothing (Fashion MNIST). Neural network training is parallelized on a NVidia graphics device using CUDA technology. There are the results for each digit illustrating the comparison of the processor and GPU operating time. For both the digits and clothing types, there is a conclusion about a 17-fold acceleration on an entry-level gaming laptop. Test samples of the same databases are used to verify the cluster construction correct-ness. For both sequential and parallel learning, it is concluded that the vectors from the test sample be-long to the correct cluster with a probability of more than 90 % in the case of handwritten digits. In ad-dition, there are calculated F-measures for each digit and type of clothing to evaluate clusters. It is shown that sequential and parallel clustering give similar results. The best values of the F-measure are obtained for the digits 0 and 1 (F-mean is 0.974), while the worst value is obtained for the digit 9 (F-mean is 0.903). For the Fashion MNIST data, the best value for the F-measure was obtained for trousers (F-average value is 0.96), and the worst value was for a shirt (F-average value is 0.34). De-spite the large variations for the F-metric values of the considered two databases, the differences in the clustering results are minimal. Thus, the maximum difference of the F-measure is about 0.01 for the MNIST and about 0.04 for the Fashion MNIST.

16. A control pixel clustering algorithm for assessing the chemical pollution impact on forest tracts from satellite photographic images [№3 за 2022 год]
Authors: Meshalkin V.P, Butusov O.B., R.R. Kantyukov , Chistyakova T.B.
Visitors: 1749
The paper proposes an original adaptive control pixel clustering algorithm for "controlled cluster anal-ysis" of satellite photographic images. The "controlled cluster analysis" algorithm is based on the premise: the possibility of using addi-tional a priori information about control pixels on satellite photographs located in different ecological zones, which allows correcting the mosaic structures and ecological zone areas taking into account ad-ditional information. The "controlled cluster analysis" algorithm differs in using additional parameters in the form of weight coefficients and control pixels, which provides more accurate binding of clustering results to ecological zones. The "controlled cluster analysis" algorithm is based on a modernized classical K-means algorithm, in which weight coefficients and control pixels are additionally introduced as param-eters. It is shown that as a result of using the "controlled cluster analysis" algorithm, the accuracy of esti-mating the size and configuration of the areas of ecological zones increases. The proposed algorithm makes it possible to calculate the total areas of ecological zones of forests more accurately, which can be proposed as a basis for assessing the degree of environmental degrada-tion and the magnitude of environmental damage to forests.

17. Computer modeling for intelligent evaluation of dynamic interaction of solids [№3 за 2022 год]
Authors: Meshkov V.V., Filatova N.N., Fedosov Yu.A.
Visitors: 2106
The paper considers the problem of integrating the computer modeling results with experimental results of dynam-ic interaction of solids using digital processing of optical and X-ray images for subsequent analysis in an infor-mation intelligent system. The novelty is in the method of assessing the differences between the results of field and computational ex-periments based on use the fuzzy logic and fuzzy sets, as well as the procedure of switching to an index scale when forming a general assessment of differences for a set of the corresponding image points. The authors pro-pose a concept and a developed algorithm for the joint cluster analysis of the results of X-ray and optical studies and the results of modeling the dynamic interaction of bodies. A computer model of solids interaction makes it possible to study the features of complex physical processes that occur when a hypersonic particle hits a spacecraft element, to substantiate the composition and design of real spacecraft elements, and to predict its damage depending on the parameters of elements and anthropogenic parti-cles within certain limits. The simulation results can become a base for forming a data bank that will expand the limited set of results of computational and experimental modeling and field experiments. In general, the computer simulation results make it possible to prepare scientifically based initial data and recommendations for developing a protection design and a protection control system for an advanced spacecraft.

18. Forecasting technological trends based on the heterogeneous data analysis [№3 за 2022 год]
Authors: Nguyen Thanh Viet , Kravets A.G.
Visitors: 2000
To achieve competitiveness, enterprises need to exploit a development strategy by forecasting promis-ing technologies using limited resources. Numerous previous studies have shown that unexpected changes in R&D and patenting of intellectual property are associated with large changes in the market value of an enterprise. In fact, there is a strong correlation between the volatility of market shares, stock prices in the early stages of high-tech enterprise development, and the period when the innova-tive technology is not yet defined. Thus, we propose that if company's share prices continue to trend upward, then the developed technologies are likely to become promising innovations in the future. This paper proposes a method for forecasting technology trends by analyzing web news to identify high-tech enterprises, predicting stock price trends for selected enterprises and analyzing the clusters of patent applications. Unlike other studies, our method advances the idea of predicting technology trends by forecasting the stock price trend using univariate and multivariate data preparation ap-proaches, and using Bayesian optimization to explore the best hyperparameters for machine and deep learning models. The study uses the developed software system for analyzing word frequency burst detection and stock price prediction. In particular, the proposed method is adopted to predict the price dynamics of Tesla and Samsung shares as case studies.

19. Information system for calculation, information accumulation and certification of phosphorite thermophysical properties [№3 за 2022 год]
Authors: Orekhov V.A. , Bobkov V.I., S.V. Panchenko
Visitors: 1994
The paper proposes creating a certificate of phosphorite properties, which allows preserving in elec-tronic form experimental temperature dependencies of the material thermal conductivity coefficient at its first heating and in the annealed state, specific true heat capacity at the first and second heating, ef-fective heat capacity at the first heating, relative linear expansion, electrical conductivity and density of phosphate material at heating. There is a developed an information system that can implement such certificate of properties. It al-lows inputting and storing data on chemical composition and thermophysical and technological proper-ties of samples in digital form, searching and processing them. This system uses client-server technolo-gy to access and process data. Calculation of thermophysical properties of phosphorites by structural models takes into account hierarchical properties of phosphorite components, their volume and mass fractions. The information for determining the component proportions is the material mineralogical composition. The information system includes a program for calculating a thermal conductivity coefficient and a specific heat capacity. It differs by taking into account mass and volume fractions of the main rock-forming minerals using interpolation of table temperature values of thermal conductivity coefficients and specific heat capacities of phosphorite, carbonate and silica to determine thermophysical proper-ties of the components at any temperature. The thermal conductivity coefficient of the material is cal-culated is based on the combined thermal conductivity model taking into account the endothermic re-action of carbonate dissociation; the specific heat capacity is calculated from the additive model. There is a developed algorithm for a program for calculating thermophysical properties by the chemical composition of phosphorites.

20. Forecasting based on the second generation artificial neural network for decision support in especially significant situations [№3 за 2022 год]
Authors: E.V. Palchevsky, Antonov V.V. , Enikeev R.R.
Visitors: 2595
Nowadays, specialized system models implemented on the basis of decision support in exceptional (emergency) situations (states) using machine learning, artificial intelligence (including using neural networks) to reproduce, predict and prevent (or minimize risk) consequences) in exceptional situations are useful and becoming increasingly popular. Floods also fall under such exceptional situations and states. Therefore, there arises the problem of early forecasting of an exceptional situation using the ex-ample of rising water levels at stationary hydrological posts in order to prevent (or minimize the risk) the transition of the system under consideration into an exceptional state (emergency situation). To solve this problem, the authors propose a decision support system for early forecasting water rise levels. It is based on a neural network (intelligent) analysis of retrospective data (code of a station-ary hydrological station / automatic station, date, water level at a stationary hydrological station / au-tomatic station, atmospheric pressure, wind speed, snow cover thickness, amount of precipitation, time and air temperature) in order to calculate the values of water levels for 5 days in advance. The artificial neural network itself is based on the freely distributed TensorFlow machine learning software library; a modified backpropagation method is used as training. Its main difference is the addition of an artificial neural network (ANN) learning rate increase factor. An analysis of the effectiveness of the proposed solution in the framework of forecasting the flood situation has shown high accuracy in calculating the forecast values of water levels: the difference be-tween the real and predicted values is 2.10 %. This will allow specialized services to carry out special-ized anti-flood measures in advance (5 days in advance). Thus, information support during special situations is an absolute (not relative) indicator of data quality that allows developing and making decisions in the framework of predicting possible critical situations and preventing the transfer of the state of the territory management system to critical situa-tions.

◄ ← Preview | 1 | 2 | 3 | Next → ►