Synchronization in Networks of Nonlinear Systems: Contraction Metric Analysis and Deep-Learning for Feedback Estimation - DYnamique, Observation et Commande des Procédés Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2022

Synchronization in Networks of Nonlinear Systems: Contraction Metric Analysis and Deep-Learning for Feedback Estimation

Mattia Giaccagli
  • Fonction : Auteur
Samuele Zoboli
  • Fonction : Auteur
Vincent Andrieu
Giacomo Casadei
  • Fonction : Auteur
  • PersonId : 1140122

Résumé

In this work, we consider the problem of global exponential synchronization of a network of identical inputaffine nonlinear time-varying systems. To this end, we tackle the problem with incremental stability tools. We propose sufficient metric-based conditions to design a distributed diffusive coupling feedback law in two frameworks. First, we consider the Euclidean scenario, where the network is assumed to be connected. Second, we tackle the Riemannian framework by assuming the presence of a leader for an undirected network. This allows considering more general systems with significant nonlinearities. In both scenarios, we propose two different control laws: a full-state feedback and a static output feedback controller. Then, we apply our design to several cases-studies. To conclude, we propose an algorithm based on deep neural networks (DNNs) to practically implement such controllers.
Fichier principal
Vignette du fichier
main_TAC_Euclidean.pdf (1.56 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03801100 , version 1 (04-11-2022)
hal-03801100 , version 2 (05-05-2023)

Identifiants

  • HAL Id : hal-03801100 , version 1

Citer

Mattia Giaccagli, Samuele Zoboli, Daniele Astolfi, Vincent Andrieu, Giacomo Casadei. Synchronization in Networks of Nonlinear Systems: Contraction Metric Analysis and Deep-Learning for Feedback Estimation. 2022. ⟨hal-03801100v1⟩
179 Consultations
276 Téléchargements

Partager

Gmail Facebook X LinkedIn More