Control-relevant neural networks for feedforward control with preview

Applied to an industrial flatbed printer

Journal article (2024)

Authors

Leontine Aarnoudse Eindhoven University of Technology
Johan Kon Eindhoven University of Technology
Wataru Ohnishi University of Tokyo
Maurice Poot Eindhoven University of Technology
Paul Tacx Eindhoven University of Technology
Nard Strijbosch Eindhoven University of Technology
T.A.E. Oomen Eindhoven University of Technology, Team Jan-Willem van Wingerden - Mechanical, Maritime and Materials Engineering
Research Group
Team Jan-Willem van Wingerden (Mechanical, Maritime and Materials Engineering) (TU Delft)
Copyright: © 2024 Leontine Aarnoudse, Johan Kon, Wataru Ohnishi, Maurice Poot, Paul Tacx, Nard Strijbosch, T.A.E. Oomen
DOI: https://doi.org/10.1016/j.ifacsc.2024.100241
More Info
expand_more

Published Date

2024

Language

English

Faculty

Mechanical, Maritime and Materials Engineering

Department

Delft Center for Systems and Control

Research Group

Team Jan-Willem van Wingerden

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.