A Right Coprime Factorization of Neural State Space Models

Jan Dimon Bendtsen

doi:10.1109/ISDA.2007.4389605

A Right Coprime Factorization of Neural State Space Models

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

Abstract

In recent years, various methods for identification of nonlinear systems in closed loop using open-loop approaches have received considerable attention. However, these methods rely on differentially coprime factorizations of the nonlinear plants, which can be difficult to compute in practice. To address this issue, this paper presents various technical results leading up to explicit formulae for right coprime factorizations of neural state space models, i.e., nonlinear system models represented in state space using neural networks, which satisfy a Bezout identity.

Original language	English
Title of host publication	Seventh International Conference on Intelligent Systems Design and Applications
Number of pages	7
Publisher	IEEE Computer Society Press
Publication date	2007
Pages	177
ISBN (Print)	0-7695-2976-3, 978-0-7695-2976-9
DOIs	https://doi.org/10.1109/ISDA.2007.4389605
Publication status	Published - 2007
Event	International Conference on Intelligent Systems Design and Applications - Rio de Janeiro, Brazil Duration: 20 Oct 2007 → 24 Oct 2007 Conference number: 7

Conference

Conference	International Conference on Intelligent Systems Design and Applications
Number	7
Country/Territory	Brazil
City	Rio de Janeiro
Period	20/10/2007 → 24/10/2007

Keywords

Youla-Kucera Parametrization
Multilayer perceptrons
State space models

Access to Document

10.1109/ISDA.2007.4389605

AUB Link

Search for the material in Aalborg University Library's search engine

Cite this

@inproceedings{281b8dd09f4011dc8188000ea68e967b,

title = "A Right Coprime Factorization of Neural State Space Models",

abstract = "In recent years, various methods for identification of nonlinear systems in closed loop using open-loop approaches have received considerable attention. However, these methods rely on differentially coprime factorizations of the nonlinear plants, which can be difficult to compute in practice. To address this issue, this paper presents various technical results leading up to explicit formulae for right coprime factorizations of neural state space models, i.e., nonlinear system models represented in state space using neural networks, which satisfy a Bezout identity. ",

keywords = "Youla-Kucera Parametrization, Multilayer perceptrons, State space models",

author = "Bendtsen, {Jan Dimon}",

year = "2007",

doi = "10.1109/ISDA.2007.4389605",

language = "English",

isbn = "0-7695-2976-3",

pages = "177",

booktitle = "Seventh International Conference on Intelligent Systems Design and Applications",

publisher = "IEEE Computer Society Press",

address = "United States",

note = "International Conference on Intelligent Systems Design and Applications ; Conference date: 20-10-2007 Through 24-10-2007",

}

TY - GEN

T1 - A Right Coprime Factorization of Neural State Space Models

AU - Bendtsen, Jan Dimon

N1 - Conference code: 7

PY - 2007

Y1 - 2007

N2 - In recent years, various methods for identification of nonlinear systems in closed loop using open-loop approaches have received considerable attention. However, these methods rely on differentially coprime factorizations of the nonlinear plants, which can be difficult to compute in practice. To address this issue, this paper presents various technical results leading up to explicit formulae for right coprime factorizations of neural state space models, i.e., nonlinear system models represented in state space using neural networks, which satisfy a Bezout identity.

AB - In recent years, various methods for identification of nonlinear systems in closed loop using open-loop approaches have received considerable attention. However, these methods rely on differentially coprime factorizations of the nonlinear plants, which can be difficult to compute in practice. To address this issue, this paper presents various technical results leading up to explicit formulae for right coprime factorizations of neural state space models, i.e., nonlinear system models represented in state space using neural networks, which satisfy a Bezout identity.

KW - Youla-Kucera Parametrization

KW - Multilayer perceptrons

KW - State space models

U2 - 10.1109/ISDA.2007.4389605

DO - 10.1109/ISDA.2007.4389605

M3 - Article in proceeding

SN - 0-7695-2976-3

SN - 978-0-7695-2976-9

SP - 177

BT - Seventh International Conference on Intelligent Systems Design and Applications

PB - IEEE Computer Society Press

T2 - International Conference on Intelligent Systems Design and Applications

Y2 - 20 October 2007 through 24 October 2007

ER -

A Right Coprime Factorization of Neural State Space Models

Abstract

Conference

Keywords

Access to Document

AUB Link

Fingerprint

Cite this