ISSN 1991-2927
 

ACP № 1 (55) 2019

'Automation of Control processes' # 3 (37) 2014

Contents
AUTOMATED CONTROL SYSTEMS

Gennady Petrovich Tokmakov, FRPC OJSC ‘RPA ‘Mars’, Doctor of Engineering; graduated from the Faculty of Radio Engineering of Ulyanovsk Polytechnic Institute; Chief Staff Scientist of FRPC OJSC ‘RPA ‘Mars’; Professor at the Department of Computer Science of Ulyanovsk State Technical University; an author of a monograph, text-books, articles, and inventions in the field of development of data models and artificial intelligence systems. [e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. ]G. Tokmakov

Representation and Processing of Information Resources in Automated System Control Functions 37_1.pdf

The article examines issues of representation and processing of information resources of an automated system and issues of formalization of up-to-date automated system control function components. This formalization is intended to eliminate replication of almost identical software modules that ensure data access through its unification. The issue under study is rather big, and its solution is offered to be discussed in a series of articles. In this article, first of the series, we performed common analysis and formal description of control function components at three levels: data, application and user interface processing. While analyzing, we find out formalized and nonformalized control function components, and we formulate a statement to solve the problem of unification of software modules ensuring access to information resource data. The following articles of this series will have data and application formalization implemented at all control function levels. Based on this formalization, we will develop a concept data model ensuring high-level information resource representation in terms of the field area. The procedure component of this model will ensure unification of processes to generate call expressions to perform information resource data access.

Control function, information resources, data control level, application level, representation level, databases, metadata bases, application unification.

Maria Ivanovna Stoliarova, ‘NII ‘Rubin’, JSC, graduated from the Faculty of Multi-Channel Telecommunications at the Bonch-Bruevich Saint-Petersburg State University of Telecommunications; a leading engineer at ‘NII ‘Rubin’, JSC; an author of articles in the field of building, analysis, and control of special-purpose telecommunication networks. [e-mail: inforubin@rubinspb.ru]M. Stoliarova

Transport Network Structure Reconstruction of a Special Purpose Telecommunication System 37_2.pdf

The article is devoted to mathematical modeling and methodology of the transport network rational structure formation given its reconstruction for use in the interests of the officials of a spe-cial purpose telecommunication system. The task of forming the transport network structure (TNS) occurs at the stage of design (planning) and operational-technical management of communication in a special purpose telecommunication system and is characterized by high dimension, by a presence in it logically independent component parts. In the article it is presented as interrelated sub-problems: TNS synthesis (topological and streaming); TNS reconstruction; estimations of external destabilizing factors (EDF) influence on a TNS; allocations of core and line resources in TNS; es-timations of time and material resources, required for the formed TNS realization. It is offered the formal description of each sub-problems and methodical facilities for their decision. It is shown that realization of the offered model and methodology in the special software of automated communication control systems complexes of facilities of special purpose communications provides TNS forming with the specified characteristics and allows quickly generate variants of its reconstruction in managing the restoration of the transport network functioning.

Transport network, synthesis, reconstruction, topological structure, streaming structure.

Konstantin Alekseevich Smirnov, ‘NII ‘Rubin’ JSC, graduated from the Faculty of Radio Communications at Military Communications University; a leading research officer of the ‘NII ‘Rubin’ JSC of Saint-Petersburg; an author of articles in the field of control of special-purpose radio nets. [e-mail: smirnov@rubin-spb.ru]K. Smirnov

About the Frequency-terrirorial Plannig of the Special Purpouse Trunked Radio 37_3.pdf

A special purpose mobile communication networks planning includes stages of frequency and territorial planning and calculation of trunking radio communication network capacity, influencing on subscribers service quality in communication networks. Frequency-territorial planning of a special purpose mobile communication networks provides a choice: a network structure (configuration), a parking lot of mobile base stations, a type, height and orientation of aerials, distribution of frequencies between base stations. At calculation a network capacity, the estimated number of the mobile subscribers, necessary number of transceivers on base stations and demanded timeliness of service with the necessary quality is considered. The analysis and synthesis of a special purpose mobile communication networks structure are rather complex problems. The decomposition method is often used when solving such problems. Components of synthesis are problems of calculation of a mobile communication networks cover zone and a problem of calculation of service zones. When solving these problems, mobile subscribers can receive estimated areas of service with the set quality. After the solution of a synthesis problem, it is necessary to carry out the analysis of the network project, using more exact calculation models for distribution of radio waves, and, if possible, to carry out tool measurements in the mobile radio communication network expansion area. The following stage is modification of network synthesis primary option. At stages of synthesis and analysis, it is necessary to use optimization methods. Considering mobile communication networks, you need to take into account four global optimization object: a mobile communication network - costs of expansion and operation, a radio channel -radio waves passing from a base station to a subscriber and backwards, availability of services - a possibility to get a set of network services at any subscriber’s location within the designated territory, quality of service - a possibility to get the declared services with the due quality in any place of the designated territory.

Frequency-territorial planning, trunked radio, base stations.

Mikhail Sergeevich Fomin, ‘NIIIT’, OJSC, Post-graduate Student of Tver State Technical University; graduated from Tver State Technical University; has master degree in ‘Informatics and Computer Engineering’; a software engineer at ‘NIIIT’, OJSC; an author of articles in the field of prediction and reliability increasing of engineering systems. [e-mail: nuclearmike@yandex.ru]M. Fomin

Simulation Model of Spta Evaluation for Automated Management Systems 37_4.pdf

The article considers the expediency of improving existing techniques for evaluating and calculating the optimal stocks of SPTA kits in area of simulation method application. Currently used analytical models have some serious limitations and assumptions. For example, the failures flow of product parts are described only by an exponential distribution. This idealization leads to a serious distortion of the real flow of spare parts requests in general and to errors in calculations. As an alternative to existing methods this paper offers a simulation model intended for evaluation and calculation of stocks in SPTA kits and in two-level SPTA systems. The logic of the main subsystems simulation model is considered when using different replenishment strategies. The article also considers the method of generating product parts failure moments with different distribution laws. Results of the comparative analysis of suggested simulation model with analytical approaches are given. The developed simulation model is implemented in software. The software is intended for design automation of SPTA kits and for evaluation of already designed SPTA kits.

Spta kit, two-level spta system, discrete-event simulation, optimal stock, replenishment strategy.

MATHEMATICAL MODELING

Aleksandr Grigorievich Tashlinskii, Ulyanovsk State Technical University, Doctor of Engineering, Professor; graduated from the Faculty of Radio Engineering at Ulyanovsk State Technical University; Deputy Head of Radio Engineering Department at Ulyanovsk State Technical University; an author of articles, monographs, and inventions in the field of digital processing of signals and images. [e-mail: tag@ulstu.ru]A. Tashlinskii,

Sergey Vasilievich Voronov, Ulyanovsk State Technical University, graduated from the Faculty of Radio Engineering of Ulyanovsk State Technical University; a post-graduate student of the same University; an author of articles in the field of estimation of parameters of geometric image deformations. [e-mail: valmedia@yandex.ru]S. Voronov

Use of Mutual Information As Objective Function of Image Parameter Quality Estimation 37_5.pdf

Mutual information is considered as the objective function of recursive image parameters’ estimation. A new approach reducing the computational cost to the image entropy estimation based on cross-validation procedure is proposed. The article offers an analysis of ways to reduce the error of mutual information gradient estimation, at this, the Parzen-window method is used for image brightness probability density estimation.

Digital image, objective function, mutual information, image entropy, cross-validation, optimization, stochastic gradient.

Anatoly Afanasyevich Gladkikh, Ulyanovsk State Technical University, Candidate of Engineering; graduated from Budyonny Military Communications Academy; finished his post-graduate studies at the same Academy; Doctoral Student at Ulyanovsk State Technical University, Professor at the Department of Telecommunications of Ulyanovsk State Technical University; an author of a monograph, text-books, articles, and patents in the field of noise-immune coding and information security. [e-mail: a.gladkikh@ulstu.ru]A. Gladkikh

Estimation of Hardware Processing Costs in a Method of Soft Algebraic Decoding of Non-binary Codes 37_6.pdf

The paper shows estimation of the implementation complexity for the algorithms of algebraic decoding of non-binary codes by giving an example of Reed-Solomon codes. Advantages of soft data processing methods over traditional models of decoders with error-correction are given. The introduced redundancy can be used at the asymptotic capability edge. The paper also suggests a suboptimal data processing algorithm to minimize the risk of decoding error. This algorithm uses a provocation method, which calls a predictable decoder reaction while correcting a deliberately erased symbol with the reliable index of soft decision. There is a significant gain in efficiency of signal-code processor in receiver due to reduction of hardware costs in computation of the error locator polynomial and its solving by regular method instead of fitting.

Algebraic non-binary code decoder, erasure, soft decoder.

Andrey Vladimirovich Ankilov, Ulyanovsk State Technical University, Candidate of Physics and Mathematics; graduated from the Faculty of Mechanics and Mathematics of Moscow State University (Branch at Ulyanovsk); Associate Professor at the Department of Higher Mathematics of Ulyanovsk State Technical University; an author of articles and monographs in the field of mathematical modeling, aerohydroelasticity. [e-mail: ankil@ulstu.ru]A. Ankilov,

Petr Aleksandrovich Velmisov, Ulyanovsk State Technical University, Doctor of Physics and Mathematics, Professor; graduated from the Faculty of Mechanics and Mathematics of Saratov State University; Head of the Department of Higher Mathematics at Ulyanovsk State Technical University; an author of articles, textbooks, and monographs in the field of mathematical modeling, aerohydroelasticity, aerohydrodynamics, differential equations. [e-mail: velmisov@ulstu.ru]P. Velmisov

Dynamics and Stability of Elastic Aileron of Aircraft Wing in Subsonic Streamline 37_7.pdf

A mathematical model of a wing with the aileron blown by a subsonic flow of the ideal gas (fluid) is offered. It is supposed that the wing is absolutely rigid, and the aileron is elastic. Dynamics and dynamic stability of the aileron are investigated. The model is described by a related system of partial differential equations for two unknown functions: the potential of velocity of gas coming on the wing and deformations of the elastic aileron. Based on the methods of the theory of functions of complex variable, the potential of velocity is expelled from the system of equations and the solution of a problem of aerohydroelasticity is consolidated to a research of the integro-differential equation containing only an unknown function of deformation of the elastic aileron. It is supposed that thickness of the elastic aileron is a variable that results in a system of equations with variable coefficients. A study of stability is performed on basis of creation of positively certain functionality corresponding to the received partial integro-differential equation. The stability conditions imposing restrictions on velocity of the incoming flow, thickness, flexural rigidity of the aileron and on other parameters of the mechanical system are received. The solution of the specified integro-differential equation for function of deformation of an element relies on basis of the Galerkin method with carrying out a numerical experiment.

Aerohydroelasticity, stability, dynamics, elastic element, wing, aileron, subsonic flow.

Andrey Vladimirovich Ankilov, Ulyanovsk State Technical University, Candidate of Physics and Mathematics; graduated from the Faculty of Mechanics and Mathematics of Moscow State University (Branch at Ulyanovsk); Associate Professor at the Department of Higher Mathematics of Ulyanovsk State Technical University; an author of articles and monographs in the field of mathematical modeling, aerohydroelasticity. [e-mail: ankil@ulstu.ru]A. Ankilov,

Petr Aleksandrovich Velmisov, Ulyanovsk State Technical University, Doctor of Physics and Mathematics, Professor; graduated from the Faculty of Mechanics and Mathematics of Saratov State University; Head of the Department of Higher Mathematics at Ulyanovsk State Technical University; an author of articles, textbooks, and monographs in the field of mathematical modeling, aerohydroelasticity, aerohydrodynamics, and differential equations. [e-mail: velmisov@ulstu.ru]P. Velmisov,

Yulia Alexandrovna Tamarova, Ulyanovsk Instrument Manufacturing Design Bureau JSC, graduated from the Faculty of Mechanics and Mathematics of Ulyanovsk State University; Software engineer of Ulyanovsk Instrument Manufacturing Design Bureau JSC; an author of articles in the field of mathematical modeling, aerohydroelasticity, aerohydrodynamics. [e-mail: kazakovaua@mail.ru]Y. Tamarova

The Dynamical Stability of an Elastic Element of the Flow Channel 37_8.pdf

A mathematical model of a device pertaining to the vibration equipment which is intended for intensification of technological processes, for example, stirring process, is offered. Operation of such devices is based on fluctuations of elastic elements blown by a flow of gas or fluid. The dynamic stability of an elastic element placed on one of the walls of the channel where a subsonic flow of gas or fluid blows (in a model of ideal compressible medium) is investigated. The model is described by a related system of partial differential equations for two unknown functions: the potential of velocity of gas or fluid and deformation of the elastic element. The problem is investigated in a linear statement corresponding to small perturbations of the flow in the channel and small deformations of the elastic element. Determination of the stability of the elastic body corresponds to a concept of stability of dynamical systems by Lyapunov. Based on building of the mixed functional, sufficient conditions of stability are obtained. Conditions impose limitations on the speed of the uniform gas flow, compressed (tensile) efforts of the element, elastic element stiffness and other parameters of the mechanical system. Examples of building stability areas for concrete parameters of the mechanical system are given.

Aerohydroelasticity, stability, dynamics, channel, elastic plate, deformation, subsonic flow.

Sergei Anatolevich Kashkirov, Deputy Head of the Design Engineering Department of Sosny R&D Company, Post-graduate Student; graduated from the Dimitrovgrad Institute of Technology, Management, and Design at Ulyanovsk State Technical University; Deputy Head of the Design Engineering Department of Sosny R&D Company; an author of articles in the field of analysis of alternating structure mechanisms. [e-mail: ksa.sosny@gmail.com]S. Kashkirov,

Vladimir Kuzmich Manzhosov, Ulyanovsk State Technical University, Doctor of Engineering, Professor; graduated from the Machine-Building Faculty of Frunze Polytechnic Institute; Head of the Department of Theoretical and Applied Mechanics at Ulyanovsk State Technical University; an author of articles, monographs, and inventions in the field of dynamics of machines, modeling of shock processes. [e-mail: v.manjosov@ulstu.ru]V. Manzhosov

The Wave Model of Hauling Member (locomotive), Elastic Rod, and Transported Object Motion  37_9.pdf

A wave model was developed to simulate motion of the rod rigidly connected to a driving member (locomotive) and towing a driven member. The motion of the driving member is set cinematically. The motion of the rod cross-sections is described by a wave equation. The equation is solved by the traveling wave method. The functions of forward and backward waves on different motion intervals are determined from the conditions of their development in the rod cross-sections interfacing the driving member and the transported object. Characteristics of the forward and backward waves are linear functions. Their view in the coordinate plane allows preparation of the field of wave states for the mechanical system. The nature and time of the wave state experienced by a random cross-section of the rod on any motion interval can be assessed by the vertical slice method for the field of wave states. The computing scheme for solving motion equations is based on representation of the rod as a multitude of successively connected sections, as well as on the pattern of the forward wave development in the cross-section х = 0, the pattern of the forward and backward wave development on the “rod end - driving member (locomotive)” boundary, and the properties of the forward and backward wave functions to maintain their parameters during propagation throughout a homogeneous section at the velocity of sound. The wave model of the rod motion allows taking into account the rod distributed mass, analyzing the wave processes that determine the nature of the motion of the transported object, and determining the unilateral constraint point.

Rod, wave equation, traveling wave method, strain wave, rod cross-section rate, longitudinal strain in rod cross-sections, field of wave states, computing scheme.

ARTIFICIAL INTELLIGENCE

Sergey Aleksandrovich Ageev, Military Academy of Communications, Candidate of Engineering, Associate Professor, Advanced Doctoral Student at the Military Academy of Communications (St. Petersburg); graduated from the Faculty of Radio-Engineering of Ulyanovsk Polytechnic Institute; specializes in the field of telecommunications system design; an author of articles and patents in the field of data-transfer system [e-mail: serg123_61@mail.r]S. Ageev,

Igor Borisovich Saenko, St. Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences, Doctor of Engineering, Professor; graduated from the Military Academy of Communications (St. Petersburg); completed his post-graduate and advanced doctoral programs at the same Academy; Professor at the St. Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences; specializes in the field of creation and development of information-management systems; an author of monographs, articles, and patents in the same field. [e-mail: ibsaen@mail.ru]I. Saenko,

Yuriy Petrovich Egorov, Federal Research-and-Production Center Open Joint Stock Company ‘Research-and-Production Association ‘Mars’, Doctor of Engineering, Professor; graduated from the Faculty of Radio-Engineering of the Admiral Makarov Leningrad Higher Marine School of Engineering; Chief Staff Scientist at Federal Research-and-Production Center Open Joint Stock Company ‘Research-and-Production Association ‘Mars’, Ulyanovsk; specializes in the field of macro-design of large-scale information-management systems; an author of monographs, articles, and patents in the field of computer-aided troop C2 systems design. [e-mail: yupe@mail.ru]Y. Egorov,

Anatoliy Afanasevich Gladkikh, Ulyanovsk State Technical University, Candidate of Engineering; Professor at the Department of Telecommunications of Ulyanovsk State Technical University; graduated from the Marshal Budenny Military Academy of Communications; completed his post-graduate studies at the same Academy; specializes in the field of signal-code sequence synthesis and antinoise coding in telecommunication systems; an author of articles and patents in this subject area [e-mail: a_gladkikh@mail.ru]A. Gladkikh,

Aleksandr Valentinovich Bogdanov, Research-and-Development Center at the Marshal Budenny Military Academy of Communications, Candidate of Military Sciences, Associate Professor, Deputy Head of the Research-and-Development Center at the Marshal Budenny Military Academy of Communications; graduated from Military Academy of Communications; completed his post-graduate studies at the same Academy; specializes in the field of creation and development of information-management systems for communications control. [e-mail: bog-saha@yandex.ru]A. Bogdanov

Intelligent Hierarchical Information Security Risk Management in Protected Special-purpose Multiservice Networks 37_10.pdf

The main approaches to the elaboration of intellectual methods and algorithms synthesized on their base for the assessment and risks management of the information security for protected multi-service networks are considered. The priority for improving such network structures is the efficiency enhancement of the management cycle, the management reliability and the expediency of the responses to the external destructive impact on the network. It is shown that the diversity, the heterogeneity, the incompleteness, the high dimensionality and the inaccuracy of the input data used in the network security management tasks predetermine the necessity of the use of means and methods of the artificial intelligence, in particular, the construction of the prediction analysis generated by the expert, which are formalized as fuzzy sets. Within the framework of this model the hierarchical interaction management tasks are developed and proved. The conceptual provisions of intellectual information security risk management network are provided; the intellectual multi-agent framework for the risk assessment of the network threats as well as the algorithms of their functioning are developed. The study results of the numerical mathematical modeling of the functioning of the intellectual multi agent relating the assessment of the threats to the information network security network are presented.

Telematic network services, tmn model, intelligent management, fuzzy inference, knowledge base, linguistic variable.

Sergey Grigorevich Cherny, Kerch State Marine Technological University, Candidate of Engineering; graduated from Kherson National Technical University; Assistant Professor at the Department of Electrical Equipment of Ships and Industrial Automation of Kerch State Marine Technological University; an author of publications in the field of expert systems. [e-mail: sergiiblack@gmail.com]S. Cherny,

Vladimir Alekseevich Dorovskoy, Kerch State Marine Technological University, Doctor of Engineering; graduated from Kryvyi Rih Mining University; Professor at the Department of Electrical Equipment of Ships and Industrial Automation of Kerch State Marine Technological University; an author of more than 100 articles and 5 monographs. [e-mail: dora1943@mail.ru]V. Dorovskoy

The Optimization Model of Decision-making Fuzzy Processes for Maritime Equipment Diagnostics 37_11.pdf

The article presents principles of fuzzy processes optimization of decision-making for sea-bed mining equipment diagnostic tests. There were developed models for the optimization of multicriteria problems in decision-making with fuzzy objectives which are described by means of the following elements: set of admissible alternates, set of all possible results, alternative evaluation, a criterion function, establishing the connection with alternatives and their evaluations, membership functions. Some rules for the uniform management of all metadata on data sources, target schemas, manipulation, scripts, etc. are provided, that use repository-based database management systems. The integrated data of procedure analysis are demonstrated by the example of the data tables. The model of indicators and factors in the form of a visualized way projected in the Curve Fitting Toolbox is represented. Methods for the isolation of particular criteria of individual subsystems that allow to consider, with the necessary level of detail, the nature of linkages and interactions of the individual parameters of detail, the nature of correlations and interactions of the individual complex deep-water control system parameters.

Optimization, decision-making, diagnostics, minerals, multicriteriality.

COMPUTER-AIDED ENGINEERING

Viktor Nikolaevich Negoda, Ulyanovsk State Technical University, Doctor of Engineering; graduated from the Faculty of Radio-Engineering of Ulyanovsk Polytechnic Institute; Professor at the Department of Computer Science of Ulyanovsk State Technical University; an author of articles in the field of computer-aided design of embedded systems; area of expertise - computer-aided design of technical systems with logical control. [e-mail: nvn@ulstu.ru]V. Negoda,

Artem Anatolevich Fedotov, Ulyanovsk State Technical University, Master of Computer Science; graduated from the Faculty of Information Systems and Technologies at Ulyanovsk State Technical University (UlSTU); an author of articles in the field of high performance computation; area of expertise - efficient computational procedure implementation on GPGPU. [e-mail: ]A. Fedotov

The Application of Gpgpu Technology in Computer-aided Control Systems Design 37_12.pdf

The article examines the computer-aided design of control system implementation by using GPGPU technology. Features of design process are analyzed, analytic and experimental procedure of decomposition of the CAD functions for the purpose of maximizing utilization of GPGPU is offered. The modified design process is based on the well-known process named COMET, created specifically for designing real-time systems. Application of GPGPU technology focuses on analytical modeling stage. Creation of software models to perform dynamic simulation of control systems is proposed by mapping related modelling functions on an array of processor cores. We construct a general logic-algebraic model of such a map. As an example of the proposed approach to the analytical modeling of the active use of GPGPU technology we consider one of the most complex challenges of designing a control system of DC furnaces - dynamic modeling of heat and mass transfer in the batch furnace based on grid model. We present the results of the experiments performed using CUDA technology.

Gpgpu technology, computer-aided design of control system, parallelization of cad procedures.

Viktor Nikolaevich Negoda, Ulyanovsk State Technical University, Doctor of Engineering; graduated from the Faculty of Radio-Engineering of Ulyanovsk Polytechnic Institute; Professor at the Department of Computer Science of Ulyanovsk State Technical University; an author of articles in the field of computer-aided design of embedded systems; area of expertise - computer-aided design of technical systems with logical control. [e-mail: nvn@ulstu.ru]V. Negoda

The Unification of Design Solutions At Automata-based Programming of Logic Control Systems   37_13.pdf

The unified set of design solutions are proposed for the efficient automata-based programming of the logic control systems. This set includes such viewing of machine state, transition and output of formal finite state machine model, that take real-time requirements into consideration while creating distributed control systems. Unification allows to establish design patterns and create reusable classes of software objects and functions. Different application aspects of the unified design solutions to the logic control system’s source code generation are considered, together with generation of various service programs for the logic control system.

Distributed control systems, logic control system design, automata-based programming, design patterns, automatic code generation.

Vadim Viktorinovich Shishkin, Ulyanovsk State Technical University, Candidate of Engineering, Associate Professor; graduated from the Faculty of Radio-Engineering at Ulyanovsk State Polytechnical Institute; Professor at the Measuring-Computing Complexes Department of Ulyanovsk State Technical University, Dean of the Faculty of Information Systems and Technologies at UlSTU, an author of articles in the field of automated design of industrial products and intellectual data analysis. [e-mail: shvv@ulstu.ru]V. Shishkin,

Denis Igorevich Stenyushkin, Ulyanovsk State Technical University, graduated from the Faculty of Information Systems and Technologies at Ulyanovsk State Technical University; Post-Graduate Student at the Measuring-Computing Complexes Department of Ulyanovsk State Technical University, an author of articles in the field of automated design of industrial products and intellectual data analysis. [e-mail: denisstenyushkin@yandex.ru]D. Stenyushkin,

Alexander Vyacheslavovich Mikheev, Ulyanovsk State Technical University, graduated from the Faculty of Information Systems and Technologies at Ulyanovsk State Technical University; Post-Graduate Student at the Measuring-Computing Complexes Department of Ulyanovsk State Technical University, an author of articles in the field of measuring-computing system design and intellectual data analysis. [e-mail: sr.alex.anderson@gmail.com]A. Mikheev

Automated Design of Artificial Neural Network Based Classifiers 37_14.pdf

The article describes a system of models and methods for automated design of neural network-based classifiers. The system bases on the closed design cycle with feedback in form of analysis of gathered experience. The system’s goal is decrease of design process’ time costs. Parameters space is utilized as a formal system for experience representation. Parameters space is a space of points that represent sets of classifiers’ parameters. The space is constructed from parameters, which values are being specified through whole classifier’s life cycle from specification up to operation. This formalism allows to formulate the experience and analysis operations performed over it in terms of sets and operations over them. This results into representation improvement and automation costs decrease. Experience analysis bases on comparison of parameters space points’ projections on dedicated directions. It allows to detect classifier design solutions that are convenient to reuse. These solutions include classifier architectures, architectural parameters and neuron input weights. Using the detected parameters values as initial values in a new classifier’s training and setting up process lead to time costs decrease of 15%.

Artificial neural network-based classifiers, classifiers design, parameters space, closed design cycle.

© FRPC JSC 'RPA 'Mars', 2009-2018 The web-site runs on Joomla!