Kernel-based learning control for iteration-varying tasks applied to a printer with friction

Maurice Poot, Jim Portegies, Tom Oomen

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

4 Citaten (Scopus)
7 Downloads (Pure)

Samenvatting

Feedforward control is essential for achieving high accuracy for nonlinear mechatronic systems. The aim of this paper is to develop a kernel-based iterative learning control (ILC) approach that enables the specification of parameters through suitable kernels. The developed kernel-based iterative learning control (KILC) framework employs basis functions to facilitate task flexibility and nonlinear and non-causal feedforward as function of the reference signal. Experimental results on a printer motion system subject to nonlinear friction demonstrate that the developed framework is capable of achieving improved performance for systems with non-minimum phase and higher-order dynamics compared to preexisting feedforward methods.

Originele taal-2Engels
Titel2021 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)
UitgeverijInstitute of Electrical and Electronics Engineers
Pagina's1052-1057
Aantal pagina's6
ISBN van elektronische versie9781665441391
DOI's
StatusGepubliceerd - 24 aug. 2021
Evenement2021 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2021 - Delft, Nederland
Duur: 12 jul. 202116 jul. 2021

Congres

Congres2021 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2021
Land/RegioNederland
StadDelft
Periode12/07/2116/07/21

Bibliografische nota

Publisher Copyright:
© 2021 IEEE.

Financiering

The work is supported by ASM Pacific Technology, Beuningen, The Netherlands.

Vingerafdruk

Duik in de onderzoeksthema's van 'Kernel-based learning control for iteration-varying tasks applied to a printer with friction'. Samen vormen ze een unieke vingerafdruk.

Citeer dit