CASPIAN JOURNAL
MANAGEMENT AND HIGH TECHNOLOGIES
ANALYSIS OF CHARACTERISTICS OF CONDITIONAL GAUSSIAN STATIONARY PROCESSES IN TASK SIMULATION SYSTEMS
Article retracted 05.02.2018
 |
Read | Yakunin Paul S., Yagudaev Gennady G., Satyshev Sergey N., Kotov Andrei A., Zhigarev Ruslan G. ANALYSIS OF CHARACTERISTICS OF CONDITIONAL GAUSSIAN STATIONARY PROCESSES IN TASK SIMULATION SYSTEMS // Caspian journal : management and high technologies. — 2011. — №3. — pp. 85-93. |
Yakunin Paul S. - Ph.D., Chief, Moscow Regional Bank (MOSOBLBANK), 3 Salsola st., Moscow, 109028, Russia, kafedra@asu.madi.ru.
Yagudaev Gennady G. - Ph.D., North Caucasus branch of the Moscow Automobile and Road Technical University (MADI), 64 Leningradsky av., Moscow, 125319, Russia, kafedra@asu.madi.ru.
Satyshev Sergey N. - Ph.D., Moscow Automobile and Road Technical University (MADI), 64 Leningradsky av., Moscow, 125319, Russia, kafedra@asu.madi.ru.
Kotov Andrei A. - Ph.D., Moscow Automobile and Road State Technical University (MADI), 64 Leningradsky av., Moscow, 125319, Russia, kafedra@asu.madi.ru.
Zhigarev Ruslan G. - graduate student, Moscow Automobile and Road Technical University (MADI), 64 Leningradsky av., Moscow, 125319, Russia, ruslan.zhigarev@mail.ru.
The article proposes the use of conventional non-stationary Gaussian process for the analytical study of the convergence process simulation, and, in particular, the convergence of algorithms in search optimization of simulation models. This process is based on the theorem on normal correlation and extrapolation is regarded as a stationary Gaussian process with a given history. The autocorrelation function of a stationary process is a mixture of two exponentials. The main objective of the study is to find dependency characteristics sredneintiegralnyh estimates obtained by non-stationary process from the initial conditions and simulation parameters of the autocorrelation function of the original stationary process. It is shown that in this class of random processes is modeling various types of monotonic trends possible, but with strong autocorrelation and distant initial conditions, and also non-monotonic trend. Dispersion of conditionally non-stationary process is characterized by a nonmonotonic behavior. At the initial stage, the variance is small because of low values of the process changes over a short period of time. Then, the variance increases, but there comes a time when the ergodic properties of the process and begin to dominate the dispersion tends to zero. It also shows that the trend has a tight estimate mean integral character, contains a significant systematic error at the initial stage and depends on the correlation of the process, the initial values and the duration of the simulation.
Key words: imitation,modeling,Gaussian processes,the mean integral evaluation,variance,trend,nonstationary processes,autocorrelation,convergence,optimization.
Причина ретракции: материал статьи в основном заимствован из докторской диссертации В.М. Черненького «Процессно-ориентированная концепция системного моделирования АСУ» (защищена в 2010 г. в г. Москве).
Решение о ретракции принято редакционной коллегией журнала «Прикаспийский журнал: управление и высокие технологии» 05.02.2018 г.