Control-relevant neural networks for feedforward control with preview: Applied to an industrial flatbed printer

L.I.M. Aarnoudse; J.J. Kon; Wataru Ohnishi; M.M. Poot; P.J.M.M. Tacx; Nard W.A. Strijbosch; Tom A.E. Oomen

doi:10.1016/j.ifacsc.2024.100241

Control-relevant neural networks for feedforward control with preview: Applied to an industrial flatbed printer

L.I.M. Aarnoudse (Corresponding author)
, J.J. Kon
, Wataru Ohnishi
, M.M. Poot
, P.J.M.M. Tacx
, Nard W.A. Strijbosch
, Tom A.E. Oomen

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

8 Citations (Scopus)

441 Downloads (Pure)

Abstract

The performance of feedforward control depends strongly on its ability to compensate for reproducible disturbances. The aim of this paper is to develop a systematic framework for artificial neural networks (ANN) for feedforward control. The method involves three aspects: a new criterion that emphasizes the closed-loop control objective, inclusion of preview to deal with delays and non-minimum phase dynamics, and enabling the use of an iterative learning algorithm to generate training data in view of addressing generalization errors. The approach is illustrated through simulations and experiments on an industrial flatbed printer.

Original language	English
Article number	100241
Journal	IFAC Journal of Systems and Control
Volume	27
DOIs	https://doi.org/10.1016/j.ifacsc.2024.100241
Publication status	Published - Mar 2024

Bibliographical note

Publisher Copyright:
© 2024 The Author(s)

Funding

This work is part of the research programme VIDI with project number 15698 , which is (partly) financed by the NWO .

Funders
Nederlandse Organisatie voor Wetenschappelijk Onderzoek

Keywords

Feedforward control
Neural networks
Iterative learning control

Access to Document

10.1016/j.ifacsc.2024.100241

2024-01_final_author_versionFinal author version , 1.13 MBLicence: CC BY-NC-ND
pdfFinal published version, 826 KBLicence: CC BY

Cite this

@article{49b14566400a4be5986f5e4f3f00d603,

title = "Control-relevant neural networks for feedforward control with preview: Applied to an industrial flatbed printer",

abstract = "The performance of feedforward control depends strongly on its ability to compensate for reproducible disturbances. The aim of this paper is to develop a systematic framework for artificial neural networks (ANN) for feedforward control. The method involves three aspects: a new criterion that emphasizes the closed-loop control objective, inclusion of preview to deal with delays and non-minimum phase dynamics, and enabling the use of an iterative learning algorithm to generate training data in view of addressing generalization errors. The approach is illustrated through simulations and experiments on an industrial flatbed printer.",

keywords = "Feedforward control, Neural networks, Iterative learning control",

author = "L.I.M. Aarnoudse and J.J. Kon and Wataru Ohnishi and M.M. Poot and P.J.M.M. Tacx and Strijbosch, \{Nard W.A.\} and Oomen, \{Tom A.E.\}",

note = "Publisher Copyright: {\textcopyright} 2024 The Author(s)",

year = "2024",

month = mar,

doi = "10.1016/j.ifacsc.2024.100241",

language = "English",

volume = "27",

journal = "IFAC Journal of Systems and Control",

issn = "2468-6018",

publisher = "Elsevier",

}

TY - JOUR

T1 - Control-relevant neural networks for feedforward control with preview

T2 - Applied to an industrial flatbed printer

AU - Aarnoudse, L.I.M.

AU - Kon, J.J.

AU - Ohnishi, Wataru

AU - Poot, M.M.

AU - Tacx, P.J.M.M.

AU - Strijbosch, Nard W.A.

AU - Oomen, Tom A.E.

PY - 2024/3

Y1 - 2024/3

N2 - The performance of feedforward control depends strongly on its ability to compensate for reproducible disturbances. The aim of this paper is to develop a systematic framework for artificial neural networks (ANN) for feedforward control. The method involves three aspects: a new criterion that emphasizes the closed-loop control objective, inclusion of preview to deal with delays and non-minimum phase dynamics, and enabling the use of an iterative learning algorithm to generate training data in view of addressing generalization errors. The approach is illustrated through simulations and experiments on an industrial flatbed printer.

AB - The performance of feedforward control depends strongly on its ability to compensate for reproducible disturbances. The aim of this paper is to develop a systematic framework for artificial neural networks (ANN) for feedforward control. The method involves three aspects: a new criterion that emphasizes the closed-loop control objective, inclusion of preview to deal with delays and non-minimum phase dynamics, and enabling the use of an iterative learning algorithm to generate training data in view of addressing generalization errors. The approach is illustrated through simulations and experiments on an industrial flatbed printer.

KW - Feedforward control

KW - Neural networks

KW - Iterative learning control

UR - https://www.scopus.com/pages/publications/85182569175

U2 - 10.1016/j.ifacsc.2024.100241

DO - 10.1016/j.ifacsc.2024.100241

M3 - Article

AN - SCOPUS:85182569175

SN - 2468-6018

VL - 27

JO - IFAC Journal of Systems and Control

JF - IFAC Journal of Systems and Control

M1 - 100241

ER -

Control-relevant neural networks for feedforward control with preview: Applied to an industrial flatbed printer

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this