| Texto completo | |
| Autor(es): |
de Oliveira Jr, Cesar Magno Leite
;
Saraiva, Antonio Mauro
;
Delbem, Alexandre Claudio Botazzo
;
Velho, Haroldo Fraga de Campos
;
Lemos, Geronimo Gallarreta Zubiaurre
;
Harter, Fabricio Pereira
Número total de Autores: 6
|
| Tipo de documento: | Artigo Científico |
| Fonte: | JOURNAL OF COMPUTATIONAL SCIENCE; v. 90, p. 11-pg., 2025-08-01. |
| Resumo | |
Data assimilation is an important process to compute the best initial condition for a computational prediction system, combining a previous prediction (background) with observation. The result from this procedure is the computed analysis. A cellular neural network (Cell-NN) is applied as a data assimilation (DA) method. The Cell-NN is also employed to integrate dynamic systems in time. Different Cell-NN configurations are developed for the DA process and as an integration scheme. The Lorenz system under a chaotic dynamical regime is used for testing with Cell-NN. Data assimilation with the 3D variational (3D-Var) method is also implemented for comparison. Cell-NN belongs to the class of unsupervised neural networks. The performance for computing the analysis by Cell-NN presented a similar error magnitude to the 3D-Var technique. (AU) | |
| Processo FAPESP: | 19/07665-4 - Centro de Inteligência Artificial |
| Beneficiário: | Fabio Gagliardi Cozman |
| Modalidade de apoio: | Auxílio à Pesquisa - Programa eScience e Data Science - Centros de Pesquisa Aplicada |