Samenvatting
Port-Hamiltonian neural networks (pHNNs) are emerging as a powerful modeling tool that integrates physical laws with deep learning techniques. However, most existing research has largely concentrated on modeling entire systems, often neglecting the thorough examination of subsystems within dynamic systems. This study presents a novel approach for identifying subsystems using pHNNs to fill this gap. By utilizing the inherent port modeling characteristics of the port-Hamiltonian systems, we have developed an algorithm that partitions port-Hamiltonian systems into distinct subsystems, allowing for more detailed analysis and modeling. To validate our approach, we conducted numerical experiments on nonlinear systems, demonstrating that it effectively captures independent subsystem dynamics.
Originele taal-2 | Engels |
---|---|
Status | Gepubliceerd - 2024 |
Evenement | 31nd Workshop of the European Research Network on System Identification - Don Orione Artigianelli, Venice, Italië Duur: 29 sep. 2024 → 2 okt. 2024 https://automatica.dei.unipd.it/32nd-ernsi-main-page/ |
Workshop
Workshop | 31nd Workshop of the European Research Network on System Identification |
---|---|
Verkorte titel | ERNSI 2024 |
Land/Regio | Italië |
Stad | Venice |
Periode | 29/09/24 → 2/10/24 |
Internet adres |