liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Output regulation of unknown linear systems using average cost reinforcement learning
Linköping University, Department of Electrical Engineering, Automatic Control. Linköping University, Faculty of Science & Engineering.
Linköping University, Department of Electrical Engineering, Automatic Control. Linköping University, Faculty of Science & Engineering.
Univ Texas Arlington, TX 76019 USA; Northeastern Univ, Peoples R China.
2019 (English)In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 110, article id 108549Article in journal (Refereed) Published
Abstract [en]

In this paper, we introduce an optimal average cost learning framework to solve output regulation problem for linear systems with unknown dynamics. Our optimal framework aims to design the controller to achieve output tracking and disturbance rejection while minimizing the average cost. We derive the Hamilton-Jacobi-Bellman (HJB) equation for the optimal average cost problem and develop a reinforcement algorithm to solve it. Our proposed algorithm is an off-policy routine which learns the optimal average cost solution completely model-free. We rigorously analyze the convergence of the proposed algorithm. Compared to previous approaches for optimal tracking controller design, we elevate the need for judicious selection of the discounting factor and the proposed algorithm can be implemented completely model-free. We support our theoretical results with a simulation example. (C) 2019 Elsevier Ltd. All rights reserved.

Place, publisher, year, edition, pages
PERGAMON-ELSEVIER SCIENCE LTD , 2019. Vol. 110, article id 108549
Keywords [en]
Output regulation; Reinforcement learning; Linear systems; Optimal control
National Category
Control Engineering
Identifiers
URN: urn:nbn:se:liu:diva-162304DOI: 10.1016/j.automatica.2019.108549ISI: 000495491900011OAI: oai:DiVA.org:liu-162304DiVA, id: diva2:1374082
Note

Funding Agencies|Vinnova Competence Center LINK-SIC; Wallenberg Artificial Intelligence, Autonomous Systems and Software Program (WASP)

Available from: 2019-11-28 Created: 2019-11-28 Last updated: 2020-02-19

Open Access in DiVA

The full text will be freely available from 2021-09-16 08:42
Available from 2021-09-16 08:42

Other links

Publisher's full text

Search in DiVA

By author/editor
Adib Yaghmaie, FarnazGunnarsson, Svante
By organisation
Automatic ControlFaculty of Science & Engineering
In the same journal
Automatica
Control Engineering

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 203 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf