Conference proceedings

 
 
Shahverdiyev Azizaga Khanbaba oglu (RU), 
Mandrik Ilya Emmanuilovich (RU), 
Panahov Geylani Minhaj oglu (AZ), 
Abbasov Eldar Mekhti oglu (AZ), 
Aliyev Hashim Mamed Jafar oglu (RU) 

---

DEEP LEARNING AND NONLINEAR RECURRENT ANALYSIS FOR 
CONTROL OF THE BEHAVIOR OF A COMPLEX SYSTEM. 
 
Bahram Ismailov 
Department of “Instrumental Engineering” 
ASOIU, Baku, AZ1010, Azerbaijan Republic 
E-mail: 
[e-mail] 
ORCID ID: 0000-0002-5815-1253 
Abstract
.
The presented report is devoted to studies of the behavior of complex systems in the framework of the Open 
System. During the research, analytical and numerical simulations of the processes of interactions between systems were used, an 
integrated approach to assessing their behavior by visualizing the dynamics was used. The presented structure and research algorithm 
covers dynamics control processes, using the example of multidimensional chaotic systems. The stages of modeling the processes of 
managing the dynamics of systems are presented in the form of time series with Areas of Interest of the researcher, the construction of 
Recurrence Plots, Finite Time Lyapunov Exponent, as well as calculations of basic information indicators such as Fractional dimension, 
Tsallis entropy, Kaplan-Yorke dimension, Poincare return time, Lyapunov exponent and others. To generalize the obtained palette of 
knowledge for making decisions on the assessment and control of the dynamics of a complex system, the Deep Learning research 
algorithm is proposed.