Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems

Zican Li; Tao Wu; Jing Na; Jun Zhao; Guanbin Gao; Guido Herrmann

doi:10.1109/ICMIC.2018.8529962

Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems

Zican Li, Tao Wu, Jing Na, Jun Zhao, Guanbin Gao, Guido Herrmann

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

2 Citations (Scopus)

367 Downloads (Pure)

Abstract

In this paper, we propose a novel method to solve the optimal output-feedback control problem of continuous-time (CT) linear systems based on a data-driven based reinforcement learning (RL). An output-feedback Riccati equation is first derived by further tailoring its counterpart of state-feedback optimal control. Then, based on this modified Riccati equation, we further derive an output Lyapunov function, where only the system output rather than the unknown state is involved. This allows to obtain the optimal output-feedback gain based on the measured output only. Then, an online data-driven based policy iteration is suggested to obtain the feedback gain K and matrix P. Finally, a simulation example is given to prove the effectiveness of the proposed algorithm.

Original language	English
Title of host publication	2018 International Conference on Modelling, Identification and Control (ICMIC 2018)
Subtitle of host publication	Proceedings of a meeting held 2-4 July 2018, Guiyang, China
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	467-472
Number of pages	6
ISBN (Electronic)	9781538654163
ISBN (Print)	9781538654170
DOIs	https://doi.org/10.1109/ICMIC.2018.8529962
Publication status	Published - Dec 2018

Keywords

Optimal control
Output-feedback control
Data-driven
Policy iteration
Riccati equation

Access to Document

10.1109/ICMIC.2018.8529962

Full-text PDF (accepted author manuscript)
This is the author accepted manuscript (AAM). The final published version (version of record) is available online via IEEE at https://ieeexplore.ieee.org/document/8529962. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 503 KB

Handle.net

Persistent link

Cite this

Li, Z., Wu, T., Na, J., Zhao, J., Gao, G., & Herrmann, G. (2018). Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems. In 2018 International Conference on Modelling, Identification and Control (ICMIC 2018): Proceedings of a meeting held 2-4 July 2018, Guiyang, China (pp. 467-472). Article 101 Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/ICMIC.2018.8529962

@inproceedings{e09f9eaa4e544392b32049daedb8f46f,

title = "Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems",

abstract = "In this paper, we propose a novel method to solve the optimal output-feedback control problem of continuous-time (CT) linear systems based on a data-driven based reinforcement learning (RL). An output-feedback Riccati equation is first derived by further tailoring its counterpart of state-feedback optimal control. Then, based on this modified Riccati equation, we further derive an output Lyapunov function, where only the system output rather than the unknown state is involved. This allows to obtain the optimal output-feedback gain based on the measured output only. Then, an online data-driven based policy iteration is suggested to obtain the feedback gain K and matrix P. Finally, a simulation example is given to prove the effectiveness of the proposed algorithm.",

keywords = "Optimal control, Output-feedback control, Data-driven, Policy iteration, Riccati equation",

author = "Zican Li and Tao Wu and Jing Na and Jun Zhao and Guanbin Gao and Guido Herrmann",

year = "2018",

month = dec,

doi = "10.1109/ICMIC.2018.8529962",

language = "English",

isbn = "9781538654170",

pages = "467--472",

booktitle = "2018 International Conference on Modelling, Identification and Control (ICMIC 2018)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

}

Li, Z, Wu, T, Na, J, Zhao, J, Gao, G & Herrmann, G 2018, Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems. in 2018 International Conference on Modelling, Identification and Control (ICMIC 2018): Proceedings of a meeting held 2-4 July 2018, Guiyang, China., 101, Institute of Electrical and Electronics Engineers (IEEE), pp. 467-472. https://doi.org/10.1109/ICMIC.2018.8529962

Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems. / Li, Zican; Wu, Tao; Na, Jing et al.
2018 International Conference on Modelling, Identification and Control (ICMIC 2018): Proceedings of a meeting held 2-4 July 2018, Guiyang, China. Institute of Electrical and Electronics Engineers (IEEE), 2018. p. 467-472 101.

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems

AU - Li, Zican

AU - Wu, Tao

AU - Na, Jing

AU - Zhao, Jun

AU - Gao, Guanbin

AU - Herrmann, Guido

PY - 2018/12

Y1 - 2018/12

N2 - In this paper, we propose a novel method to solve the optimal output-feedback control problem of continuous-time (CT) linear systems based on a data-driven based reinforcement learning (RL). An output-feedback Riccati equation is first derived by further tailoring its counterpart of state-feedback optimal control. Then, based on this modified Riccati equation, we further derive an output Lyapunov function, where only the system output rather than the unknown state is involved. This allows to obtain the optimal output-feedback gain based on the measured output only. Then, an online data-driven based policy iteration is suggested to obtain the feedback gain K and matrix P. Finally, a simulation example is given to prove the effectiveness of the proposed algorithm.

AB - In this paper, we propose a novel method to solve the optimal output-feedback control problem of continuous-time (CT) linear systems based on a data-driven based reinforcement learning (RL). An output-feedback Riccati equation is first derived by further tailoring its counterpart of state-feedback optimal control. Then, based on this modified Riccati equation, we further derive an output Lyapunov function, where only the system output rather than the unknown state is involved. This allows to obtain the optimal output-feedback gain based on the measured output only. Then, an online data-driven based policy iteration is suggested to obtain the feedback gain K and matrix P. Finally, a simulation example is given to prove the effectiveness of the proposed algorithm.

KW - Optimal control

KW - Output-feedback control

KW - Data-driven

KW - Policy iteration

KW - Riccati equation

U2 - 10.1109/ICMIC.2018.8529962

DO - 10.1109/ICMIC.2018.8529962

M3 - Conference Contribution (Conference Proceeding)

SN - 9781538654170

SP - 467

EP - 472

BT - 2018 International Conference on Modelling, Identification and Control (ICMIC 2018)

PB - Institute of Electrical and Electronics Engineers (IEEE)

ER -

Li Z, Wu T, Na J, Zhao J, Gao G, Herrmann G. Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems. In 2018 International Conference on Modelling, Identification and Control (ICMIC 2018): Proceedings of a meeting held 2-4 July 2018, Guiyang, China. Institute of Electrical and Electronics Engineers (IEEE). 2018. p. 467-472. 101 Epub 2018 Nov 12. doi: 10.1109/ICMIC.2018.8529962

Data-Driven Based Optimal Output-Feedback Control of Continuous-Time Systems

Abstract

Keywords

Access to Document

Handle.net

Fingerprint

Cite this