Open AccessFORMATION OF A HIGHLY INFORMATIVE DIGITAL SIGNAL IMAGE OF AN AUTOMATED CONTROL SYSTEM USING TIME-FREQUENCY REPRESENTATION AND TWO-DIMENSIONAL DIGITAL FILTERING TECHNOLOGIES
Author(s) -
Andrey N. Ragozin,
A.V. Portnov,
S.S. Lysov,
N.S. Prytkov
Publication year - 2020
Publication title -
vestnik urfo. bezopasnostʹ v informacionnoj sfere
Language(s) - English
Resource type - Journals
eISSN - 2225-5443
pISSN - 2225-5435
DOI - 10.14529/secur200308
Subject(s) - signal (programming language) , computer science , digital signal , representation (politics) , process (computing) , artificial intelligence , digital image , digital image processing , signal processing , digital filter , digital signal processing , image processing , image (mathematics) , computer vision , pattern recognition (psychology) , computer hardware , filter (signal processing) , politics , political science , law , programming language , operating system
The study examines the technology of forming a high-informative digital image of the ob-served time signal of an automated control system (ACS). Formation of the signal image in-cludes the stage of displaying the signal on the time-frequency representation, obtaining the time-frequency image of the signal, the stage of two-dimensional digital filtering, the stage of formation of high-informative components of the signal image. The resulting digital image of the observed ACS signal represents the generated high-informative input data for further solv-ing a wide range of different problems using machine learning technologies and deep neural networks in the ACS of industrial systems. A wide range of tasks using the generated digital im-age of the signal implies, for example, the analysis of process behavior and identification of anomalies in the process of ACS, the implementation of diagnostic algorithms of technical equipment in the ACS, the detection of changes in the dynamics of the process of ACS under the influence of information attacks