TY - JOUR

T1 - Цифровой алгоритм адаптивного управления контролируемыми переменными в заданном диапазоне

AU - Сотникова, Маргарита Викторовна

AU - Севостьянов, Руслан Андреевич

PY - 2025

Y1 - 2025

N2 - The article is devoted to the issues of a digital control system synthesis for an arbitrary nonlinear object, including taking into account the delay. An important feature is the absence of a reference signal, the purpose of control is to maintain the controlled variables in a given range. The problem under consideration is solved based on the control method with a predictive model and a special quality functional. It is shown that the optimal control problem at each step is reduced to a nonlinear optimization problem, while the admissible set is never empty due to the use of auxiliary variables. This approach allows changing the restrictions on the control and measured variables in real time. Another important feature is the ability to turn on and off the variables of the control object during its operation, as well as to adjust the prediction for cases where the model is not known exactly. The efficiency of the proposed approach is demonstrated using the example of experiments with a computer model of a rectification column for refining petroleum products.

AB - The article is devoted to the issues of a digital control system synthesis for an arbitrary nonlinear object, including taking into account the delay. An important feature is the absence of a reference signal, the purpose of control is to maintain the controlled variables in a given range. The problem under consideration is solved based on the control method with a predictive model and a special quality functional. It is shown that the optimal control problem at each step is reduced to a nonlinear optimization problem, while the admissible set is never empty due to the use of auxiliary variables. This approach allows changing the restrictions on the control and measured variables in real time. Another important feature is the ability to turn on and off the variables of the control object during its operation, as well as to adjust the prediction for cases where the model is not known exactly. The efficiency of the proposed approach is demonstrated using the example of experiments with a computer model of a rectification column for refining petroleum products.

KW - adaptivity

KW - control in a given range

KW - digital control

UR - https://www.mendeley.com/catalogue/5d55aa0c-a53b-35ac-8b87-306c44e05dcf/

U2 - 10.21638/spbu10.2025.111

DO - 10.21638/spbu10.2025.111

M3 - статья

VL - 21

SP - 151

EP - 168

JO - ВЕСТНИК САНКТ-ПЕТЕРБУРГСКОГО УНИВЕРСИТЕТА. ПРИКЛАДНАЯ МАТЕМАТИКА. ИНФОРМАТИКА. ПРОЦЕССЫ УПРАВЛЕНИЯ

JF - ВЕСТНИК САНКТ-ПЕТЕРБУРГСКОГО УНИВЕРСИТЕТА. ПРИКЛАДНАЯ МАТЕМАТИКА. ИНФОРМАТИКА. ПРОЦЕССЫ УПРАВЛЕНИЯ

SN - 1811-9905

IS - 1

ER -