Learning the dynamic of clusters of vertices in a graph for District Heating Network simplification
Résumé
District Heating Networks (DHNs) provide very efficient and flexible solutions to produce and supply heat energy for local uses but are computationally expensive both to optimize and simulate. Leveraging the formulation of a DHN as series of graphs with time series signals on its vertices, the objective of this work is to reduce such computational costs by aggregating identified vertices. We investigate recurrent neural network model to learn and mimic the temporal dynamic of the signals of aggregated vertices.
Fichier principal
Rodrigue et al. - Learning the dynamic of clusters of vertices in a .pdf (350.91 Ko)
Télécharger le fichier
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|