Online Data-Enabled Predictive Control

Stefanos Baros; Chin-Yao Chang; Gabriel Colon-Reyes; Andrey Bernstein

doi:10.1016/j.automatica.2021.109926

Online Data-Enabled Predictive Control

Stefanos Baros, Chin-Yao Chang, Gabriel Colon-Reyes, Andrey Bernstein

Power Systems Engineering

University of California, Berkeley

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

19 Scopus Citations

Abstract

We develop an online data-enabled predictive (ODeePC) control method for optimal control of unknown systems, building on the recently proposed DeePC (Coulson et al., 2019). Our proposed ODeePC method leverages a primal–dual algorithm with real-time measurement feedback to iteratively compute the corresponding real-time optimal control policy as system conditions change. The proposed ODeePC conceptual-wise resembles standard adaptive system identification and model predictive control (MPC), but it provides a new alternative for the standard methods. ODeePC is enabled by computationally efficient methods that exploit the special structure of the Hankel matrices in the context of DeePC with Fast Fourier Transform (FFT) and primal–dual algorithm We provide theoretical guarantees regarding the asymptotic behavior of ODeePC, and we demonstrate its performance through numerical examples.

Original language	American English
Article number	109926
Number of pages	9
Journal	Automatica
Volume	138
DOIs	https://doi.org/10.1016/j.automatica.2021.109926 https://doi.org/10.1016/j.automatica.2021.109926
State	Published - 2022

Bibliographical note

Publisher Copyright:
© 2021

NREL Publication Number

NREL/JA-5D00-76265

Keywords

Data-driven control
Model predictive control
Online optimization

Access to Document

Cite this

@article{7cda7ec7bd59461abef7cd6e49a67e89,

title = "Online Data-Enabled Predictive Control",

abstract = "We develop an online data-enabled predictive (ODeePC) control method for optimal control of unknown systems, building on the recently proposed DeePC (Coulson et al., 2019). Our proposed ODeePC method leverages a primal–dual algorithm with real-time measurement feedback to iteratively compute the corresponding real-time optimal control policy as system conditions change. The proposed ODeePC conceptual-wise resembles standard adaptive system identification and model predictive control (MPC), but it provides a new alternative for the standard methods. ODeePC is enabled by computationally efficient methods that exploit the special structure of the Hankel matrices in the context of DeePC with Fast Fourier Transform (FFT) and primal–dual algorithm We provide theoretical guarantees regarding the asymptotic behavior of ODeePC, and we demonstrate its performance through numerical examples.",

keywords = "Data-driven control, Model predictive control, Online optimization",

author = "Stefanos Baros and Chin-Yao Chang and Gabriel Colon-Reyes and Andrey Bernstein",

note = "Publisher Copyright: {\textcopyright} 2021",

year = "2022",

doi = "10.1016/j.automatica.2021.109926",

language = "American English",

volume = "138",

journal = "Automatica",

issn = "0005-1098",

publisher = "Elsevier Ltd.",

}

TY - JOUR

T1 - Online Data-Enabled Predictive Control

AU - Baros, Stefanos

AU - Chang, Chin-Yao

AU - Colon-Reyes, Gabriel

AU - Bernstein, Andrey

PY - 2022

Y1 - 2022

N2 - We develop an online data-enabled predictive (ODeePC) control method for optimal control of unknown systems, building on the recently proposed DeePC (Coulson et al., 2019). Our proposed ODeePC method leverages a primal–dual algorithm with real-time measurement feedback to iteratively compute the corresponding real-time optimal control policy as system conditions change. The proposed ODeePC conceptual-wise resembles standard adaptive system identification and model predictive control (MPC), but it provides a new alternative for the standard methods. ODeePC is enabled by computationally efficient methods that exploit the special structure of the Hankel matrices in the context of DeePC with Fast Fourier Transform (FFT) and primal–dual algorithm We provide theoretical guarantees regarding the asymptotic behavior of ODeePC, and we demonstrate its performance through numerical examples.

AB - We develop an online data-enabled predictive (ODeePC) control method for optimal control of unknown systems, building on the recently proposed DeePC (Coulson et al., 2019). Our proposed ODeePC method leverages a primal–dual algorithm with real-time measurement feedback to iteratively compute the corresponding real-time optimal control policy as system conditions change. The proposed ODeePC conceptual-wise resembles standard adaptive system identification and model predictive control (MPC), but it provides a new alternative for the standard methods. ODeePC is enabled by computationally efficient methods that exploit the special structure of the Hankel matrices in the context of DeePC with Fast Fourier Transform (FFT) and primal–dual algorithm We provide theoretical guarantees regarding the asymptotic behavior of ODeePC, and we demonstrate its performance through numerical examples.

KW - Data-driven control

KW - Model predictive control

KW - Online optimization

UR - http://www.scopus.com/inward/record.url?scp=85123871700&partnerID=8YFLogxK

U2 - 10.1016/j.automatica.2021.109926

DO - 10.1016/j.automatica.2021.109926

M3 - Article

AN - SCOPUS:85123871700

SN - 0005-1098

VL - 138

JO - Automatica

JF - Automatica

M1 - 109926

ER -

Online Data-Enabled Predictive Control

Abstract

Bibliographical note

NREL Publication Number

Keywords

Access to Document

Other files and links

Fingerprint

Cite this