A MATLAB toolbox for training and implementing physics-guided neural network-based feedforward controllers

M. Bolderman (Corresponding author), Mircea Lazar (Corresponding author), Hans Butler (Corresponding author)

Onderzoeksoutput: Bijdrage aan tijdschriftCongresartikelpeer review

41 Downloads (Pure)

Samenvatting

Physics-guided neural networks (PGNNs) enable accurate identification of inverse system dynamics by effectively embedding a known physical model within a neural network (NN), and thereby achieve high performance when implemented as feedforward controllers. However, training PGNNs using existing NN toolboxes is complicated. Therefore, this paper presents a MATLAB toolbox that systematically implements, trains, and validates PGNNs. Dedicated functions implement recent results that have been proposed in literature, i.e., we ensure that the PGNN converges to a value of the cost function that is strictly upperbounded by the value obtained when using only the physical model, while also imposing a form of graceful degradation when the trained PGNN is used on data that was not present in the training data. The toolbox is available at:https://github.com/mbolderman/PGNN-Toolbox/
Originele taal-2Engels
Pagina's (van-tot)4068-4073
Aantal pagina's6
TijdschriftIFAC-PapersOnLine
Volume56
Nummer van het tijdschrift2
DOI's
StatusGepubliceerd - 1 jul. 2023
Evenement22nd World Congress of the International Federation of Automatic Control (IFAC 2023 World Congress) - Yokohama, Japan
Duur: 9 jul. 202314 jul. 2023
Congresnummer: 22
https://www.ifac2023.org/

Financiering

This work is part of the research programme with project number 17973, which is (partly) financed by the Dutch Research Council (NWO).

FinanciersFinanciernummer
Nederlandse Organisatie voor Wetenschappelijk Onderzoek

    Vingerafdruk

    Duik in de onderzoeksthema's van 'A MATLAB toolbox for training and implementing physics-guided neural network-based feedforward controllers'. Samen vormen ze een unieke vingerafdruk.

    Citeer dit