DOI

Problems of convergence and stability of Bayesian estimates in the identification of stochastic control systems are considered. The informational measure of the mismatch between the estimated distribution and the estimate is the main apparatus for establishing the fact of convergence. The choice of a priori distribution of parameters is not always obvious. The Kullback-Leibler information number is taken as such measure. The convergence of the estimates of the transition function of the process to the non-stationary transition function is established in this paper. The problem of synthesis of optimal strategies for dynamic systems in which there is no part of the main information needed for constructing the optimal control is also considered. It is assumed that the system contains at least one unknown parameter belonging to some parameter space. Therefore, the class of control systems considered in the article is the class of parametric adaptive systems.

Язык оригиналаанглийский
Название основной публикацииComputational Science and Its Applications- ICCSA 2019 - 19th International Conference, Proceedings
РедакторыBeniamino Murgante, Osvaldo Gervasi, Elena Stankova, Vladimir Korkhov, Sanjay Misra, Carmelo Torre, Eufemia Tarantino, David Taniar, Ana Maria A.C. Rocha, Bernady O. Apduhan
ИздательSpringer Nature
Страницы691-701
Число страниц11
ISBN (печатное издание)9783030243043
DOI
СостояниеОпубликовано - 1 янв 2019
Событие19th International Conference on Computational Science and Its Applications, ICCSA 2019 - Saint Petersburg, Российская Федерация
Продолжительность: 1 июл 20194 июл 2019
Номер конференции: 19

Серия публикаций

НазваниеLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Том11622 LNCS
ISSN (печатное издание)0302-9743
ISSN (электронное издание)1611-3349

конференция

конференция19th International Conference on Computational Science and Its Applications, ICCSA 2019
Сокращенное названиеICCSA 2019
Страна/TерриторияРоссийская Федерация
ГородSaint Petersburg
Период1/07/194/07/19

    Предметные области Scopus

  • Теоретические компьютерные науки
  • Компьютерные науки (все)

ID: 43689589