Результаты исследований: Научные публикации в периодических изданиях › статья › Рецензирование
СИНТЕЗ ОПТИМАЛЬНЫХ ЛИНЕЙНЫХ РЕГУЛЯТОРОВ С ОГРАНИЧЕНИЕМ НА ОБЛАСТЬ РАСПОЛОЖЕНИЯ КОРНЕЙ. / Сотникова, Маргарита Викторовна; Томилова, Анастасия Сергеевна; Лепихин, Тимур Андреевич.
в: Журнал Средневолжского математического общества, Том 28, № 2, 08.07.2026, стр. 47-61.Результаты исследований: Научные публикации в периодических изданиях › статья › Рецензирование
}
TY - JOUR
T1 - СИНТЕЗ ОПТИМАЛЬНЫХ ЛИНЕЙНЫХ РЕГУЛЯТОРОВ С ОГРАНИЧЕНИЕМ НА ОБЛАСТЬ РАСПОЛОЖЕНИЯ КОРНЕЙ
AU - Сотникова, Маргарита Викторовна
AU - Томилова, Анастасия Сергеевна
AU - Лепихин, Тимур Андреевич
PY - 2026/7/8
Y1 - 2026/7/8
N2 - Currently, the synthesis of LQR controllers (i.e., linear quadratic regulators) is one of the most significant and widely used method for designing optimal stabilizing controllers. This method guarantees the asymptotic stability of a closed-loop system, which means that the roots of its characteristic polynomial are placed in the open left half-plane of the complex plane. This paper proposes an extension for the capabilities of classical LQR synthesis. Namely, the considered problem is designing an optimal linear controller that minimizes a quadratic performance functional with an additional constraint on the root domain of the closed-loop system’s characteristic polynomial. A sector in the left half-plane of the complex plane is introduced as such a region, which makes it possible to influence such performance indices of transient processes in a closed-loop system as the transient time and oscillation. The root domain is parameterized using Euclidean space vectors. It is shown that the parametric synthesis problem of obtaining an optimal controller, taking into account modal constraints, reduces to unconstrained minimization. A computational algorithm for obtaining the optimal controller is developed, which is based on solving a linear Lyapunov matrix equation at each iteration of the optimization method, but does not require solving the nonlinear Riccati equation. Searching for weight matrices of a quadratic functional in the synthesis of an LQR controller without constraints, corresponding to the optimal controller in a problem with modal constraints, is considered. The obtained theoretical results are illustrated with a practical example of vessel course control.
AB - Currently, the synthesis of LQR controllers (i.e., linear quadratic regulators) is one of the most significant and widely used method for designing optimal stabilizing controllers. This method guarantees the asymptotic stability of a closed-loop system, which means that the roots of its characteristic polynomial are placed in the open left half-plane of the complex plane. This paper proposes an extension for the capabilities of classical LQR synthesis. Namely, the considered problem is designing an optimal linear controller that minimizes a quadratic performance functional with an additional constraint on the root domain of the closed-loop system’s characteristic polynomial. A sector in the left half-plane of the complex plane is introduced as such a region, which makes it possible to influence such performance indices of transient processes in a closed-loop system as the transient time and oscillation. The root domain is parameterized using Euclidean space vectors. It is shown that the parametric synthesis problem of obtaining an optimal controller, taking into account modal constraints, reduces to unconstrained minimization. A computational algorithm for obtaining the optimal controller is developed, which is based on solving a linear Lyapunov matrix equation at each iteration of the optimization method, but does not require solving the nonlinear Riccati equation. Searching for weight matrices of a quadratic functional in the synthesis of an LQR controller without constraints, corresponding to the optimal controller in a problem with modal constraints, is considered. The obtained theoretical results are illustrated with a practical example of vessel course control.
KW - Lyapunov matrix equation
KW - linear controller
KW - marine vessel
KW - modal constraints
KW - quadratic functional
KW - unconstrained optimization
UR - https://www.elibrary.ru/item.asp?id=91818515
UR - https://www.mendeley.com/catalogue/1c98be57-6a9c-3330-9cbe-6c9ce3f893db/
U2 - 10.15507/2079-6900.28.202602.47-61
DO - 10.15507/2079-6900.28.202602.47-61
M3 - статья
VL - 28
SP - 47
EP - 61
JO - Zhurnal Srednevolzhskogo Matematicheskogo Obshchestva
JF - Zhurnal Srednevolzhskogo Matematicheskogo Obshchestva
SN - 2079-6900
IS - 2
ER -
ID: 159510543