Online Linear Quadratic Tracking with Regret Guarantees

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Karapetyan, Aren, Bolliger, Diego, Tsiamis, Anastasios, Balta, Efe C., Lygeros, John
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929547121786880
author Karapetyan, Aren
Bolliger, Diego
Tsiamis, Anastasios
Balta, Efe C.
Lygeros, John
author_facet Karapetyan, Aren
Bolliger, Diego
Tsiamis, Anastasios
Balta, Efe C.
Lygeros, John
contents Online learning algorithms for dynamical systems provide finite time guarantees for control in the presence of sequentially revealed cost functions. We pose the classical linear quadratic tracking problem in the framework of online optimization where the time-varying reference state is unknown a priori and is revealed after the applied control input. We show the equivalence of this problem to the control of linear systems subject to adversarial disturbances and propose a novel online gradient descent based algorithm to achieve efficient tracking in finite time. We provide a dynamic regret upper bound scaling linearly with the path length of the reference trajectory and a numerical example to corroborate the theoretical guarantees.
format Preprint
id arxiv_https___arxiv_org_abs_2303_10260
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Online Linear Quadratic Tracking with Regret Guarantees
Karapetyan, Aren
Bolliger, Diego
Tsiamis, Anastasios
Balta, Efe C.
Lygeros, John
Systems and Control
Online learning algorithms for dynamical systems provide finite time guarantees for control in the presence of sequentially revealed cost functions. We pose the classical linear quadratic tracking problem in the framework of online optimization where the time-varying reference state is unknown a priori and is revealed after the applied control input. We show the equivalence of this problem to the control of linear systems subject to adversarial disturbances and propose a novel online gradient descent based algorithm to achieve efficient tracking in finite time. We provide a dynamic regret upper bound scaling linearly with the path length of the reference trajectory and a numerical example to corroborate the theoretical guarantees.
title Online Linear Quadratic Tracking with Regret Guarantees
topic Systems and Control
url https://arxiv.org/abs/2303.10260