Distributed adaptive dynamic programming for data-driven optimal control

Wentao Tang; Prodromos Daoutidis

doi:10.1016/j.sysconle.2018.08.002

Distributed adaptive dynamic programming for data-driven optimal control

Wentao Tang, Prodromos Daoutidis

Chemical Engineering and Materials Science

Research output: Contribution to journal › Article › peer-review

26 Scopus citations

Abstract

Adaptive dynamic programming (ADP), as an important optimal control technique, can be exploited in the setting of data-driven control based on an approximate regression-based solution of the Hamilton–Jacobi–Bellman (HJB) equations. Distributed optimization algorithms, which are extensively studied in statistics and machine learning, have not yet been applied to the solution of data-driven ADP problems. In this work, we identify the data-driven ADP problem as a consensus optimization problem for nonlinear affine systems, and apply the alternating direction method of multipliers (ADMM) and its accelerated variants for its solution. For the input-constrained optimal control problem, we define a combined optimal primal–dual function to develop a data-based version of the input-constrained HJB equation.

Original language	English (US)
Pages (from-to)	36-43
Number of pages	8
Journal	Systems and Control Letters
Volume	120
DOIs	https://doi.org/10.1016/j.sysconle.2018.08.002
State	Published - Oct 2018

Bibliographical note

Funding Information:
This work was supported by National Science Foundation, USA (NSF-CBET).

Publisher Copyright:
© 2018 Elsevier B.V.

Keywords

Adaptive dynamic programming
Data-driven control
Distributed optimization
Process control

Publisher link

10.1016/j.sysconle.2018.08.002

Find It

Find It at U of M Libraries

Cite this

@article{c465664dc8544e5cb5ce1bd1cc3400a2,

title = "Distributed adaptive dynamic programming for data-driven optimal control",

abstract = "Adaptive dynamic programming (ADP), as an important optimal control technique, can be exploited in the setting of data-driven control based on an approximate regression-based solution of the Hamilton–Jacobi–Bellman (HJB) equations. Distributed optimization algorithms, which are extensively studied in statistics and machine learning, have not yet been applied to the solution of data-driven ADP problems. In this work, we identify the data-driven ADP problem as a consensus optimization problem for nonlinear affine systems, and apply the alternating direction method of multipliers (ADMM) and its accelerated variants for its solution. For the input-constrained optimal control problem, we define a combined optimal primal–dual function to develop a data-based version of the input-constrained HJB equation.",

keywords = "Adaptive dynamic programming, Data-driven control, Distributed optimization, Process control",

author = "Wentao Tang and Prodromos Daoutidis",

note = "Funding Information: This work was supported by National Science Foundation, USA (NSF-CBET). Publisher Copyright: {\textcopyright} 2018 Elsevier B.V.",

year = "2018",

month = oct,

doi = "10.1016/j.sysconle.2018.08.002",

language = "English (US)",

volume = "120",

pages = "36--43",

journal = "Systems and Control Letters",

issn = "0167-6911",

publisher = "Elsevier",

}

TY - JOUR

T1 - Distributed adaptive dynamic programming for data-driven optimal control

AU - Tang, Wentao

AU - Daoutidis, Prodromos

PY - 2018/10

Y1 - 2018/10

N2 - Adaptive dynamic programming (ADP), as an important optimal control technique, can be exploited in the setting of data-driven control based on an approximate regression-based solution of the Hamilton–Jacobi–Bellman (HJB) equations. Distributed optimization algorithms, which are extensively studied in statistics and machine learning, have not yet been applied to the solution of data-driven ADP problems. In this work, we identify the data-driven ADP problem as a consensus optimization problem for nonlinear affine systems, and apply the alternating direction method of multipliers (ADMM) and its accelerated variants for its solution. For the input-constrained optimal control problem, we define a combined optimal primal–dual function to develop a data-based version of the input-constrained HJB equation.

AB - Adaptive dynamic programming (ADP), as an important optimal control technique, can be exploited in the setting of data-driven control based on an approximate regression-based solution of the Hamilton–Jacobi–Bellman (HJB) equations. Distributed optimization algorithms, which are extensively studied in statistics and machine learning, have not yet been applied to the solution of data-driven ADP problems. In this work, we identify the data-driven ADP problem as a consensus optimization problem for nonlinear affine systems, and apply the alternating direction method of multipliers (ADMM) and its accelerated variants for its solution. For the input-constrained optimal control problem, we define a combined optimal primal–dual function to develop a data-based version of the input-constrained HJB equation.

KW - Adaptive dynamic programming

KW - Data-driven control

KW - Distributed optimization

KW - Process control

UR - http://www.scopus.com/inward/record.url?scp=85053123123&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053123123&partnerID=8YFLogxK

U2 - 10.1016/j.sysconle.2018.08.002

DO - 10.1016/j.sysconle.2018.08.002

M3 - Article

AN - SCOPUS:85053123123

SN - 0167-6911

VL - 120

SP - 36

EP - 43

JO - Systems and Control Letters

JF - Systems and Control Letters

ER -

Distributed adaptive dynamic programming for data-driven optimal control

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this