Identification from data with periodically missing output samples

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Autoren

Organisationseinheiten

Externe Organisationen

  • Institució Catalana de Recerca i Estudis Avançats (ICREA)
  • International Centre for Numerical Methods in Engineering
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Aufsatznummer111869
Seitenumfang5
FachzeitschriftAutomatica
Jahrgang169
Frühes Online-Datum22 Aug. 2024
PublikationsstatusElektronisch veröffentlicht (E-Pub) - 22 Aug. 2024

Abstract

The identification problem in case of data with missing values is challenging and currently not fully understood. For example, there are no general nonconservative identifiability results, nor provably correct data efficient methods. In this paper, we consider a special case of periodically missing output samples, where all but one output sample per period may be missing. The novel idea is to use a lifting operation that converts the original problem with missing data into an equivalent standard identification problem. The key step is the inverse transformation from the lifted to the original system, which requires computation of a matrix root. The well-posedness of the inverse transformation depends on the eigenvalues of the system. Under an assumption on the eigenvalues, which is not verifiable from the data, and a persistency of excitation-type assumption on the data, the method based on lifting recovers the data-generating system.

ASJC Scopus Sachgebiete

Zitieren

Identification from data with periodically missing output samples. / Markovsky, Ivan; Alsalti, Mohammad Salahaldeen Ahmad; Lopez Mejia, Victor Gabriel et al.
in: Automatica, Jahrgang 169, 111869, 11.2024.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Markovsky I, Alsalti MSA, Lopez Mejia VG, Müller MA. Identification from data with periodically missing output samples. Automatica. 2024 Nov;169:111869. Epub 2024 Aug 22. doi: 10.15488/15821, 10.1016/j.automatica.2024.111869
Markovsky, Ivan ; Alsalti, Mohammad Salahaldeen Ahmad ; Lopez Mejia, Victor Gabriel et al. / Identification from data with periodically missing output samples. in: Automatica. 2024 ; Jahrgang 169.
Download
@article{71bcfe8e4e434348af4d077aa428f7cd,
title = "Identification from data with periodically missing output samples",
abstract = "The identification problem in case of data with missing values is challenging and currently not fully understood. For example, there are no general nonconservative identifiability results, nor provably correct data efficient methods. In this paper, we consider a special case of periodically missing output samples, where all but one output sample per period may be missing. The novel idea is to use a lifting operation that converts the original problem with missing data into an equivalent standard identification problem. The key step is the inverse transformation from the lifted to the original system, which requires computation of a matrix root. The well-posedness of the inverse transformation depends on the eigenvalues of the system. Under an assumption on the eigenvalues, which is not verifiable from the data, and a persistency of excitation-type assumption on the data, the method based on lifting recovers the data-generating system.",
keywords = "Behavioral approach, Lifting, Missing data, System identification",
author = "Ivan Markovsky and Alsalti, {Mohammad Salahaldeen Ahmad} and {Lopez Mejia}, {Victor Gabriel} and M{\"u}ller, {Matthias A.}",
note = "Publisher Copyright: {\textcopyright} 2024 Elsevier Ltd",
year = "2024",
month = aug,
day = "22",
doi = "10.15488/15821",
language = "English",
volume = "169",
journal = "Automatica",
issn = "0005-1098",
publisher = "Elsevier Ltd.",

}

Download

TY - JOUR

T1 - Identification from data with periodically missing output samples

AU - Markovsky, Ivan

AU - Alsalti, Mohammad Salahaldeen Ahmad

AU - Lopez Mejia, Victor Gabriel

AU - Müller, Matthias A.

N1 - Publisher Copyright: © 2024 Elsevier Ltd

PY - 2024/8/22

Y1 - 2024/8/22

N2 - The identification problem in case of data with missing values is challenging and currently not fully understood. For example, there are no general nonconservative identifiability results, nor provably correct data efficient methods. In this paper, we consider a special case of periodically missing output samples, where all but one output sample per period may be missing. The novel idea is to use a lifting operation that converts the original problem with missing data into an equivalent standard identification problem. The key step is the inverse transformation from the lifted to the original system, which requires computation of a matrix root. The well-posedness of the inverse transformation depends on the eigenvalues of the system. Under an assumption on the eigenvalues, which is not verifiable from the data, and a persistency of excitation-type assumption on the data, the method based on lifting recovers the data-generating system.

AB - The identification problem in case of data with missing values is challenging and currently not fully understood. For example, there are no general nonconservative identifiability results, nor provably correct data efficient methods. In this paper, we consider a special case of periodically missing output samples, where all but one output sample per period may be missing. The novel idea is to use a lifting operation that converts the original problem with missing data into an equivalent standard identification problem. The key step is the inverse transformation from the lifted to the original system, which requires computation of a matrix root. The well-posedness of the inverse transformation depends on the eigenvalues of the system. Under an assumption on the eigenvalues, which is not verifiable from the data, and a persistency of excitation-type assumption on the data, the method based on lifting recovers the data-generating system.

KW - Behavioral approach

KW - Lifting

KW - Missing data

KW - System identification

UR - http://www.scopus.com/inward/record.url?scp=85201676283&partnerID=8YFLogxK

U2 - 10.15488/15821

DO - 10.15488/15821

M3 - Article

VL - 169

JO - Automatica

JF - Automatica

SN - 0005-1098

M1 - 111869

ER -

Von denselben Autoren