Deep model-free KKL observer: A switching approach - Archive ouverte HAL Access content directly
Conference Papers Year : 2024

Deep model-free KKL observer: A switching approach

Abstract

This paper presents a new model-free methodology to learn Kazantzis-Kravaris-Luenberger (KKL) observers for nonlinear systems. We address three major difficulties arising in observer design: the peaking phenomenon, the noise sensitivity and the trade-off between convergence speed and robustness. We formulate the learning objective as an optimization problem, strictly minimizing the error of the observer estimates, without the need of adding explicit constraints or regularization terms. We further improve the performance with a switching approach, efficiently transitioning between two observers, respectively designed for the transient phase and the asymptotic convergence. Numerical results on the Van der Pol system, the Rössler attractor and on a bioreactor illustrate the gain of the method regarding the literature, in term of performance and robustness. Code available online: https://github.com/jolindien-git/DeepKKL
Fichier principal
Vignette du fichier
2024_DeepKKL_switching.pdf (1.12 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04518181 , version 1 (23-03-2024)

Identifiers

  • HAL Id : hal-04518181 , version 1

Cite

Johan Peralez, Madiha Nadri. Deep model-free KKL observer: A switching approach. Learning for Dynamics and Control, Jul 2024, Oxford, United Kingdom. ⟨hal-04518181⟩
71 View
11 Download

Share

Gmail Facebook X LinkedIn More