ISSN 0236-235X (P)
ISSN 2311-2735 (E)

Bookmark

Next issue

1
Publication date:
16 March 2021
-->

Journal articles №4 2020

11. Experimental analysis of the accuracy and performance of varieties of YOLO architectures for computer vision problems [№4 за 2020 год]
Author: P.A. Bokov (deathhodd@gmail.com) - Volgograd State Technical University (Student); P.D. Kravchenya (kpd_@mail.ru ) - Volgograd State Technical University (Senior Lecturer);
Abstract: Unmanned vehicles are increasingly being introduced into everyday life. To achieve full autonomy when traveling, unmanned vehicles use computer vision systems, which are responsible for analyzing the status of traffic lights, signs, and other objects that can appear on the road. Today, the standard in this area is YOLOv1 architecture, however, it is already obsolete. In this regard, a computer vision sys-tem for an unmanned vehicle based on modern technologies is being developed. There is the problem of choosing a computer vision architecture that will be responsible for analyz-ing traffic. First of all, it must be fast and accurate, because road traffic changes very quickly, and the accuracy of determination directly affects the degree of involvement of passenger-drivers in the pro-cess in order to avoid emergency situations. In addition, the architecture should occupy as little compu-ting power as possible, and not waste a large number of energy resources. To investigate these issues, it was decided to carry out an experiment that would reveal the advantages and disadvantages of various YOLO architectures. Also, the data provided by different researchers is very different due to using dif-ferent equipment while training and testing of networks. This makes it impossible for data to be com-pared objectively. The paper analyzes various types of YOLOv3 architecture and its versions for low-power compu-ting systems YOLOv3-tiny, describes their advantages and disadvantages for computer vision systems. The experiments are carried out on single hardware for all analyzed architectures. Experimental re-search on the accuracy and performance of various YOLO architectures is being done. The VOC2012 dataset is used for training and testing. As a result of the research, the strengths and weaknesses of the architectures under consideration are determined and options for the further development of the tech-nology are analyzed, taking into account the growth in the power of computing systems and the emer-gence of new technological solutions.
Keywords: computer vision, software development, unmanned vehicles, neural network, yolov3
Visitors: 330

12. Simulation of the heat conduction process using cellular automata systems [№4 за 2020 год]
Authors: S.P. Bobkov (bsp@isuct.ru) - Ivanovo State University of Chemical Technology (Professor), Ph.D; E.G. Galiaskarov (bsp@isuct.ru) - Ivanovo State University of Chemical Technology (Associate Professor ), Ph.D;
Abstract: The paper concentrates on the use of discrete dynamic models as an alternative to the classical meth-ods of studying thermal processes in chemical technology. An adequate description of heat transfer phenomena is a hugely important problem, both in theoret-ical terms and from the standpoint of the practical use of thermal processes. In addition, modern teach-ing methods require the introduction of electronic textbooks, virtual laboratory workshops, simulation programs, which also need a correct description of the phenomena under study. The classical approach to modeling heat transfer in a continuous medium involves the use of heat conduction equations, in which the thermophysical characteristics of materials are usually constants. Taking into account the effect of temperature on the characteristics of materials leads to the need to study nonlinear equations, which causes significant computational difficulties. In this regard, it be-comes expedient to use fundamentally different approaches to modeling thermal conductivity, one of which is models based on systems of cellular automata. Discrete dynamic models in the form of deterministic cellular automata systems are used. In this case, a continuous medium is considered as a set of interacting elements whose behavior is completely described by local functions. The paper describes the main approaches and general methodology for the development of discrete models. The examples of cellular automata systems use for simulation of nonlinear heat transfer processes are considered, taking into account the heterogeneity of the material and the presence in the material of volumetric sources of variable power in it. The obtained data of discrete simulation are in good agreement with the results of using the classi-cal approach and do not contradict the generally accepted views adopted in the theory of thermal phe-nomena. The paper shows the features of the discrete approach in comparison with the use of partial differential equations with nonlinear coefficients, shows the advantages and disadvantages.
Keywords: volumetric heat sources, nonlinear heat conduction problems, heat transfer, cellular automata, discrete simulation
Visitors: 333

13. Method of functioning of the onboard radar system while ensuring their stealth operation on radiation [№4 за 2020 год]
Authors: A.V. Bogdanov (denjuga68@yandex.ru) - Military Academy of the Aerospace Defence (Professor ), Ph.D; D.V. Zakomoldin (denjuga68@yandex.ru) - Military Academy of the Aerospace Defence (Doctoral Student), Ph.D; S.I. Akimov (denjuga68@yandex.ru) - Military Academy of the Aerospace Defence (Adjunct );
Abstract: The survivability of a military aircraft depends to a significant extent on the survivability of its on-Board radar station, the main areas of improvement of which are, first, the application of the multi-position principle of building on-Board radars, and secondly, increasing the stealth of its work on radi-ation. This paper sets and solves the problem of developing a method that combines these areas of in-creasing survivability. The application of the multi-position principle is implemented by combining all on-Board radars into a single system controlled by the on-Board radar of the leader aircraft, which is determined in advance and serves as a point for processing radar information received from the on-Board radars of all the group's aircraft and issuing information to all the group's aircraft about the re-quired parameters of their on-Board radars. Secrecy operation of the system of airborne radar at the ra-diation detection group of enemy aircraft, equipped with electronic intelligence stations, implemented by means of reception on each side of the radar system information from the aircraft-leader about the required parameters, their on-Board radars, namely the values of the average radiation power of the transmitter, the time of coherent accumulation in the receiver and the time of irradiation of air targets and the formation of current data values of managed parameters of the onboard radar so that the differ-ence between the required and current values of the onboard radar parameters is zero. Held on Board the aircraft-leader calculations given in this paper, the results of these calculations for each on-Board radar systems as well as manage settings of each onboard radar systems allow you to control the signal-noise generated at the input of the receivers of all stations of electronic intelligence of the enemy, and thereby to ensure the secrecy of the operation of the onboard radar system for radia-tion with a given probability upon detection of a group of enemy aircraft equipped with radio engineer-ing reconnaissance.
Keywords: pulse-doppler airborne radar, fighter, radio intelligent station, stealth operation, parameter control, radiation power, coherent accumulation time, multi-position radar, system
Visitors: 329

14. Method of synthesis of adaptive radio technical monitoring system [№4 за 2020 год]
Authors: S.V. Butsev (butsev57@bk.ru) - Research Center of Central Research Institute of Air and Space Forces (Leading Researcher), Ph.D;
Abstract: Radio technical monitoring systems for various mission objectives operate in the presence of both un-certainties in the parameters of the monitored process and the uncertainty of a generalized disturbance. The author of the paper has developed a method for synthesizing algorithms for the functioning of an adaptive radio technical monitoring system, consisting of an adaptive filter and an adaptive control system. The synthesis of an adaptive filter for a monitoring input signal includes the development of a slid-ing adaptation algorithm based on a direct estimate of the filter parameters, in particular, the gain weights generated by the adaptation unit and used in the main filter of the monitoring system. The paper proposes to use a two-level structure, which includes two stages of synthesis: the main control loop (optimal controller) and the adaptation loop (adaptive controller). The optimal controller is synthesized on the basis of the principles of the theory of optimal control of stochastic processes, provided that the parameters of the control object of the radio engineering tracking system are constant and known, and external disturbances do not change (or are absent). The synthesis of the adaptive con-troller is carried out for the case of the simultaneous presence of the uncertainty of the parameters of the control object of the radio technical monitoring system and external disturbances acting on it (model of disturbances of the control object of the radio technical monitoring system), based on the re-current modification of the least-squares identification method. The proposed approach makes it possible to formalize the problem of the functioning of an adaptive radio technical monitoring system under conditions of uncertainty of generalized disturbance parame-ters. New relationships are obtained for the evaluation of gains of an adaptive filter and an adaptive regulator transmission matrix. The adaptive radio technical monitoring system developed on the basis of the proposed method en-sures efficient functioning within the formalized quality description on the basis of the indicator de-termined by the quadratic function which characterizes the accuracy of the system operation and its control costs. The paper provides some efficiency evaluation results of the synthesized adaptive radio tracking system functioning.
Keywords: radio technical monitoring system, control object, generalized disturbance, control device, adaptive filter, adaptive control system, adaptive regulator, operation algorithms
Visitors: 292

15. Development of a computational environment for the simulation of gas transmission systems regimes based on telemetry data [№4 за 2020 год]
Author: Е.А. Golubyatnikov (golubyatnikov.e@gubkin.ru ) - National University of Oil and Gas "Gubkin University" (Senior Lecturer);
Abstract: The paper discusses the problems of software systems for modeling the regimes of pipeline systems based on telemetry data for dispatch control. The author analyzes the subject area, as well as the fea-tures of the modeling software implementation. As a result, the requirements for such systems are for-mulated. The main requirements are modularity; extensibility and flexibility of integration mechanisms with enterprise information systems and calculation modules; organization of complex and autono-mous computing process; support for distributed component interactions. It is noted that the regime-modeling software operated in the gas oil and gas transportation industry today do not fully meet the stated requirements. Therefore, the paper proposes to develop a specialized distributed computing sim-ulation environment based on telemetry. The paper presents architectural solutions for the computing environment. A microservice approach was chosen as the basis for creating the architecture. According to the ap-proach, the designed system is divided into small, context-sensitive functional blocks. The author pro-poses the way to decompose the developing system into services, describes the roles and functions of each service and methods for service integration. The developed architectural solutions were tested during dispatch control of a real gas transporta-tion system. The paper presents the implementation of the developed architecture. It is integrated with the SCADA-systems of the enterprise for the exchange of telemetry data and simulation results, as well as the Vesta software for solving hydraulic modeling problems. The created software product is used by dispatching personnel on a daily basis and allows solving urgent problems of operational manage-ment: real-time modeling, forecasting the process, calculation of analytical indicators of the system’s functioning.
Keywords: gas transportation systems, dispatch control, computational environment, integrated hardware and software system, microservice software architecture, on-line modeling
Visitors: 312

16. The software for the subsystem of quality control of manufactured products using intelligent algorithms [№4 за 2020 год]
Author: Grishin E.S. (negrshn@mail.ru) - Immanuel Kant Baltic Federal University (Postgraduate Student);
Abstract: This paper solves the problem of reducing the production of welded pipes from stainless steel grades of products of inadequate quality by creating software that functions as part of the plant's process control system and provides quality control of products using intelligent algorithms. The author describes an algorithm for continuous control of product quality, the result of which is a conclusion about the quality of products. On its basis, a subsystem of continuous quality control was developed, which controls the quality of finished products, based on a database of materials and tech-nological maps of manufactured products. This subsystem was developed as part of the creation and implementation of an integrated automat-ed process control system (APCS) and a subsystem for continuous diagnostics and equipment condi-tion monitoring using intelligent algorithms based on machine learning. To implement intelligent ma-chine learning algorithms, the open-source ML.NET cross-platform modeling framework was used, which allows you to get a model based on input data and simplifies the integration of the model into a finished .NET application. If necessary, the framework allows you to train additionally or retrain the model. The subsystem of continuous diagnostics and monitoring of the state of equipment is based on the production model of knowledge representation, which in turn is based on the processing of diag-nostic rules. Diagnostic rules are developed for specific production and unit of equipment by a special-ist in this subject area. The result of the work of the subsystem for quality control of manufactured products is the control of the characteristics of technological equipment that affect the quality of products, control of the characteristics of products based on the data of production flow charts, the issuance of warnings about the tendency for the observed characteristics to leave the range of permissible values and information about incipient defects in products, associated with these characteristics. As a result of the develop-ment of software for the subsystem of quality control of manufactured products using intelligent algo-rithms, the number of products of inadequate quality has been reduced due to the early detection of de-fects (wear) of equipment.
Keywords: knowledge base, production model, interface, machine learning, ml.net
Visitors: 327

17. Development of a problem-oriented management system for the construction of geotechnological wells [№4 за 2020 год]
Author: D.N. Moldashi (moldashi@gmail.com) - National University of Scince Research Technological "MISIS" (Applicant);
Abstract: Characteristic a research object. Geotechnological well support control problem-oriented system is in-tended for remote monitoring and drilling operations dispatching control. In the system being created, control and management activities, such as drilling processes control and drilling equipment according to measuring instruments, working logs maintenance, dispatching control, analysis, and emergency sit-uations monitoring, are automated, performed by technological personnel and expedition management. The main idea a problem-oriented control system creating is to automate the collection and automation of objects information on the current state processing, control, and drilling work control, as well as emergency situations monitoring. The subject area of the problem-oriented control system can be as-signed to automation objects with distributed organizational structure, close information relations, and information, considerable volumes processing discrete character. The aim of the paper is to increase the geological exploration expeditions drilling team’s efficiency on the basis of the problem-oriented system creation for managing the geotechnological wells support, which allows implementing procedures for monitoring the drilling operations parameters at the field, maintaining automated accounting of the drilling equipment operation modes and visualizing various drilling indicators. The novelty of the research is to provide the possibility of drilling team’s production performance in-depth analysis in real-time, drilling parameters control, and solved problems visualization in techno-logical processes continuous nature conditions and drilling rigs territorial distribution. Results of the work. Geotechnological well support control problem-oriented system will allow im-plementing the measurement functions and different drilling modes parameters control and drilling equipment state in real-time. In the system, there is a possibility of drilling parameters indirect meas-urements with subsequent structural relationships calculation, analysis, and corresponding information output. Problem-oriented control system software allows in real-time to receive, process, and visualize data from sensors, to calculate drilling process parameters using mathematical models, and to notify the driller about technological parameters deviation.
Keywords: problem-oriented management system, the hierarchical distributed system, drilling operations monitoring, jobs automation, operational performance analysis
Visitors: 330

18. Optimization of lighting calculation for interior scenes for stochastic ray tracing [№4 за 2020 год]
Authors: S.V. Ershov (voloboy@gin.keldysh.ru ) - Keldysh Institute of Applied Mathematics of RAS (Senior Researcher), Ph.D; I.V. Valiev (voloboy@gin.keldysh.ru ) - Keldysh Institute of Applied Mathematics of RAS (Researcher); A.G. Voloboy (voloboy@gin.keldysh.ru ) - Keldysh Institute of Applied Mathematics of RAS (Leading Researcher), Ph.D;
Abstract: Lighting simulation and the creation of realistic images increases the efficiency of building design. The most common method used for this simulation is bidirectional Monte Carlo ray tracing. The paper discusses the improvement of the Monte Carlo ray tracing which optimizes the emitting of rays from a natural light source for interior scenes. The proposed algorithms can accelerate the con-vergence of the method, i.e. reduce the noise remaining after a certain simulation time, and thus im-prove the final image of the virtual scene. These algorithms are effective for indoor scenes illuminated externally through windows or other openings. Rays from light sources are generated so that they are directed into these windows, deliberately cutting off the rays that do not illuminate the interior. In other words, the number of rays increases for directions that contribute to the image formed by the camera. The first of the proposed algorithms require supplementary user specifications for marking win-dows, transparent doors, or other openings through which light can enter into interior rooms. The sec-ond algorithm is fully automatic. It belongs to a wide class of algorithms for finding and using the opti-mal PDF (Probability density function) in stochastic ray tracing. The second algorithm provides less acceleration for simulation but does not require action from the user, in particular, eliminating the po-tential error that can lead to underestimation of the room illumination. The paper provides the mathematical justification of the correctness of the proposed algorithms. The examples show that the proposed approach allows calculating the image with the desired quality several times faster.
Keywords: optimal pdf, monte-carlo method, Kepler orbit, lighting simulation, realistic graphics, computer graphics
Visitors: 342

19. Automatic tracking method of stem cells of the brain by MRI data [№4 за 2020 год]
Author: M.V. Shustova (m.v.shustova@gmail.com) - Program Systems Institute of RAS (Postgraduate Student);
Abstract: The use of stem cells for the treatment of serious diseases is a promising area of modern biomedical re-search. After transplantation to the places of the disease, the stem cells presumably have a restorative effect, including reducing the foci of ischemic brain damage after a stroke, but the exact mechanism of action remains unclear. There is an urgent need to develop new effective and safe approaches to treat-ing patients by stem cell transplantation. The successful use of cell therapy is possible only after a thorough study of the interaction of cells with the local microenvironment (migration paths to organs and tissues of the body, their effect on the lesion sites) and the development of algorithms for the ef-fective use of stem cells. The combination of modern methods of magnetic resonance imaging and methods of scientific vis-ualization makes it possible to build maps of migration and distribution of stem cells after transplanta-tion. The latter is extremely important for understanding the mechanisms of stem cell action in ischem-ic stroke of the brain for the preparation and planning of clinical trials. To solve the problem of track-ing stem cells, the availability of appropriate tools is required. To date, there are a large number of software tools for processing and analyzing biomedical, but they are not enough to automatically track the pathways of stem cell migration and the dynamics of changes of ischemic stroke foci. There is a need to develop new tools to solve these problems. This paper concentrates on the automatic tracking of mesenchymal stem cell’s clusters transplanted into the brain of laboratory animals affected by an ischemic stroke. The formalization of the solution is reduced to a transportation problem. The most probable movement of clusters is visualized using cog-nitive graphics. Simultaneously with stem cell tracking, a modeling of the area of the ischemic lesion is performed. The developed toolkit provides calculations related to the unification of the brain represen-tations of various individual animals with the aim of subsequent measurement of damaged areas; per-forming deep processing and cognitive visualization associated with the scientific highlighting of the areas of interest of the researcher. The modeling and measurement results were analyzed for the effect of transplanted cells on the is-chemic stroke foci. The analysis using the developed tools showed that the stem cell’s clusters really migrate to the area of ischemic lesion, and the volume of the ischemic focus decreases with time.
Keywords: homing, mesenchymal stem cells, ischemic stroke, tracking, magnetic resonance imaging, cognitive graphics, vehicle routing problem, image processing
Visitors: 339

20. Comparative analysis of DBMS for tourist social network [№4 за 2020 год]
Author: E.F. Feoktistov (freshstyler@mail.ru) - Volgograd State Technical University (Student);
Abstract: Digital technologies are widely used in all spheres of human life, including tourism. In order to book a ready-made tour no longer need to go to a travel agency; Hotel reservations abroad and air travel can also be done without leaving home. Viewing mobile applications with prices for hotels, tickets, and tours has become part of the daily life of a tourist – even when he is not going on a trip, he often opens these applications. So there was a goal to create a mobile application for tourists for daily use. The niche of the tourist social network where tourists could communicate, plan their trips, arrange a cultur-al souvenir exchange, is empty. In this regard, a tourist social network is being developed with a recommendation system based on fairly simple personal data of tourists: lists of souvenirs for exchange (what is what they want); upcom-ing and past trips; the city of residence, and nationality. There is a problem with choosing the database management system necessary for this problem. Firstly, it should be scalable, in view of the possible large influx of users from different countries. Sec-ondly, it must meet modern requirements, be reliable and fast. The paper analyzes various types of non-relational DBMS (database management systems), based on the experience of using them in other social networks. Their advantages and disadvantages are de-scribed for subsequent possible use in a tourist social network. Three graph DBMSs were also tested: Virtuoso, Neo4j, and Sesame to identify the most reliable and fastest DBMS for this development. As a result, on the basis of the data obtained, the best DBMS was revealed, which passed most of the tests with the best time results.
Keywords: software development, tourism, DBMS, socialnetwork, database
Visitors: 321

← Preview | 1 | 2 | 3 | Next →