Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities

Azita Dabiri; Kanghui He; Shengling Shi; Dingshan Sun; Jesus Lago; Bart De Schutter

doi:10.1007/978-3-031-82681-8_6

Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities

Azita Dabiri^*, Kanghui He, Shengling Shi, Dingshan Sun, Jesus Lago, Bart De Schutter

^*Corresponding author for this work

Research output: Chapter in Book/Conference proceedings/Edited volume › Chapter › Scientific › peer-review

Abstract

Learning-based control, in particularReinforcement Learning (RL) reinforcementReinforcement learning, and optimization-based control, in particular model predictive control, each have their advantages and disadvantages for online, real-timeOptimal control optimal controlOptimal control of systems with complex dynamicsDynamic. However, both approaches are highly complementary and therefore there is an increased interest in combining their advantages in an integrated approach. In this chapter, we provide an overview of recent results, challenges, and opportunities on an integrated learning-based and optimization-based control approach. We focus in particular on piecewise affine systems as they are an extension of linear systemsLinear systems that can model or approximate hybridHybrid or nonlinearNonlinearbehaviorBehavior and as they still allow for effective numerical solutionSolution approaches.

Original language	English
Title of host publication	Nonlinear and Constrained Control
Subtitle of host publication	Applications, Synergies, Challenges and Opportunities
Editors	Emanuele Garone, Ilya Kolmanovsky, Tam W. Nguyen
Publisher	Springer
Chapter	6
Pages	131-149
Number of pages	19
ISBN (Electronic)	978-3-031-82681-8
ISBN (Print)	978-3-031-82680-1, 978-3-031-82683-2
DOIs	https://doi.org/10.1007/978-3-031-82681-8_6
Publication status	Published - 2025

Publication series

Name	Lecture Notes in Control and Information Sciences
Volume	496
ISSN (Print)	0170-8643
ISSN (Electronic)	1610-7411

Bibliographical note

Green Open Access added to TU Delft Institutional Repository as part of the Taverne amendment. More information about this copyright law amendment can be found at https://www.openaccess.nl.

Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Access to Document

10.1007/978-3-031-82681-8_6

Cite this

Dabiri, A., He, K., Shi, S., Sun, D., Lago, J., & De Schutter, B. (2025). Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities. In E. Garone, I. Kolmanovsky, & T. W. Nguyen (Eds.), Nonlinear and Constrained Control : Applications, Synergies, Challenges and Opportunities (pp. 131-149). (Lecture Notes in Control and Information Sciences; Vol. 496). Springer. https://doi.org/10.1007/978-3-031-82681-8_6

Dabiri, Azita ; He, Kanghui ; Shi, Shengling et al. / Integrating Learning-Based and MPC-Based Control for PWA Systems : Challenges and Opportunities. Nonlinear and Constrained Control : Applications, Synergies, Challenges and Opportunities. editor / Emanuele Garone ; Ilya Kolmanovsky ; Tam W. Nguyen. Springer, 2025. pp. 131-149 (Lecture Notes in Control and Information Sciences).

@inbook{10f514f1eddd4c73afc9a3a5ba9a6b52,

title = "Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities",

abstract = "Learning-based control, in particularReinforcement Learning (RL) reinforcementReinforcement learning, and optimization-based control, in particular model predictive control, each have their advantages and disadvantages for online, real-timeOptimal control optimal controlOptimal control of systems with complex dynamicsDynamic. However, both approaches are highly complementary and therefore there is an increased interest in combining their advantages in an integrated approach. In this chapter, we provide an overview of recent results, challenges, and opportunities on an integrated learning-based and optimization-based control approach. We focus in particular on piecewise affine systems as they are an extension of linear systemsLinear systems that can model or approximate hybridHybrid or nonlinearNonlinearbehaviorBehavior and as they still allow for effective numerical solutionSolution approaches.",

author = "Azita Dabiri and Kanghui He and Shengling Shi and Dingshan Sun and Jesus Lago and {De Schutter}, Bart",

note = "Green Open Access added to TU Delft Institutional Repository as part of the Taverne amendment. More information about this copyright law amendment can be found at https://www.openaccess.nl. Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. ",

year = "2025",

doi = "10.1007/978-3-031-82681-8_6",

language = "English",

isbn = "978-3-031-82680-1",

series = "Lecture Notes in Control and Information Sciences",

publisher = "Springer",

pages = "131--149",

editor = "Emanuele Garone and Ilya Kolmanovsky and Nguyen, {Tam W.}",

booktitle = "Nonlinear and Constrained Control",

}

Dabiri, A , He, K , Shi, S , Sun, D, Lago, J & De Schutter, B 2025, Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities. in E Garone, I Kolmanovsky & TW Nguyen (eds), Nonlinear and Constrained Control : Applications, Synergies, Challenges and Opportunities. Lecture Notes in Control and Information Sciences, vol. 496, Springer, pp. 131-149. https://doi.org/10.1007/978-3-031-82681-8_6

Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities. / Dabiri, Azita ; He, Kanghui ; Shi, Shengling et al.
Nonlinear and Constrained Control : Applications, Synergies, Challenges and Opportunities. ed. / Emanuele Garone; Ilya Kolmanovsky; Tam W. Nguyen. Springer, 2025. p. 131-149 (Lecture Notes in Control and Information Sciences; Vol. 496).

Research output: Chapter in Book/Conference proceedings/Edited volume › Chapter › Scientific › peer-review

TY - CHAP

T1 - Integrating Learning-Based and MPC-Based Control for PWA Systems

T2 - Challenges and Opportunities

AU - Dabiri, Azita

AU - He, Kanghui

AU - Shi, Shengling

AU - Sun, Dingshan

AU - Lago, Jesus

AU - De Schutter, Bart

N1 - Green Open Access added to TU Delft Institutional Repository as part of the Taverne amendment. More information about this copyright law amendment can be found at https://www.openaccess.nl. Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2025

Y1 - 2025

N2 - Learning-based control, in particularReinforcement Learning (RL) reinforcementReinforcement learning, and optimization-based control, in particular model predictive control, each have their advantages and disadvantages for online, real-timeOptimal control optimal controlOptimal control of systems with complex dynamicsDynamic. However, both approaches are highly complementary and therefore there is an increased interest in combining their advantages in an integrated approach. In this chapter, we provide an overview of recent results, challenges, and opportunities on an integrated learning-based and optimization-based control approach. We focus in particular on piecewise affine systems as they are an extension of linear systemsLinear systems that can model or approximate hybridHybrid or nonlinearNonlinearbehaviorBehavior and as they still allow for effective numerical solutionSolution approaches.

AB - Learning-based control, in particularReinforcement Learning (RL) reinforcementReinforcement learning, and optimization-based control, in particular model predictive control, each have their advantages and disadvantages for online, real-timeOptimal control optimal controlOptimal control of systems with complex dynamicsDynamic. However, both approaches are highly complementary and therefore there is an increased interest in combining their advantages in an integrated approach. In this chapter, we provide an overview of recent results, challenges, and opportunities on an integrated learning-based and optimization-based control approach. We focus in particular on piecewise affine systems as they are an extension of linear systemsLinear systems that can model or approximate hybridHybrid or nonlinearNonlinearbehaviorBehavior and as they still allow for effective numerical solutionSolution approaches.

UR - http://www.scopus.com/inward/record.url?scp=105021333350&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-82681-8_6

DO - 10.1007/978-3-031-82681-8_6

M3 - Chapter

AN - SCOPUS:105021333350

SN - 978-3-031-82680-1

SN - 978-3-031-82683-2

T3 - Lecture Notes in Control and Information Sciences

SP - 131

EP - 149

BT - Nonlinear and Constrained Control

A2 - Garone, Emanuele

A2 - Kolmanovsky, Ilya

A2 - Nguyen, Tam W.

PB - Springer

ER -

Dabiri A , He K , Shi S , Sun D, Lago J, De Schutter B. Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities. In Garone E, Kolmanovsky I, Nguyen TW, editors, Nonlinear and Constrained Control : Applications, Synergies, Challenges and Opportunities. Springer. 2025. p. 131-149. (Lecture Notes in Control and Information Sciences). doi: 10.1007/978-3-031-82681-8_6

Integrating Learning-Based and MPC-Based Control for PWA Systems: Challenges and Opportunities

Abstract

Publication series

Bibliographical note

Access to Document

Other files and links

Embargoed Document

Fingerprint

Cite this