Redes Neurais Artificiais na Melhoria de Desempenho de Métodos de Assimilação de Dados: Filtro de Kalman
DOI:
https://doi.org/10.5540/tema.2010.011.01.0029Abstract
Assimilação de Dados é um método que combina dados de um modelo matemático e dados de observações, permitindo uma melhoria na previsão do modelo. Métodos seqüenciais ótimos são baseados em teoria de estimativa formal que minimiza os erros dos dados de acordo com a dinâmica do modelo. Métodos de assimilação de dados utilizando Redes Neurais Artificiais (RNA) vêm sendo propostos muito recentemente apresentando resultados consistentes: computacionalmente eficientes e eficazes quanto aplicação. Este trabalho apresenta uma abordagem do método de assimilação por RNA, onde aplica-se uma RNA para substituir o cálculo da inversão de matrizes de erros constante do algoritmo de assimilação baseado em filtro de Kalman. Para exemplo da aplicação desta abordagem, utilizou-se o Sistema de Lorenz e o Filtro de Kalman Estendido para obter parâmetros usados no treinamento da RNA e na comparação dos resultados.References
[1] F. Boutier, P. Courtier, Data Assimilation concepts and methods, in “Meteorological Training Course”, Reading, UK, 1998.
S.D. Conte, C. de Boor, “Elementary Numerical Analysis: an Algorithmic Approach”, McGraw-Hill, 3rd Edition, 1980.
D. Coppersmith, S. Winograd, Matrix multiplication via arithmetic progressions, Journal of Symbolic Computation, 9 (1990), 251–280.
R. Daley, “Atmospheric Data Analysis”, Cambridge University Press, 1991.
M.W. Gardner, S.R. Dorling, Artificial neural networks (the multilayer perpectron) – a review of applications in the atmospheric sciences, Atmospheric Environment, 32, No. 14–15 (1998), 2627–2636.
M.J. Goris, D.A. Gray, I.M.Y. Marcels, Reduncing the computational load of a Kalman filter, IEE Eletronics Letters, 33, No. 12 (1997), 1539–1541.
F.P. Härter, H.F. de Campos Velho, New approach to applying neural network in nonlinear dynamic model, Applied Mathematical Modelling, 32 (2008), 2621–2633. doi: 10.1016/j.apm.2007.09.006
F.P. Härter, “Redes Neurais Recorrentes Aplicadas à Assimilação de Dados em Dinâmica Não Linear", Tese de Doutorado, Computação Aplicada, CAP-INPE, São José dos Campos, SP, 2004.
F.P. Härter, H.F. de Campos Velho, Recurrent and feedforward neural networks trained with cross correlation applied to the data assimilation in chaotic dynamic, Revista Brasileira de Meteorologia, 20, No. 3 (2005), 411–420.
R.E. Kalman, A new approach to linear filtering and prediction problems, Trans. of the ASME–Journal of Basic Engineering, 82, No. D (1960), 35–45.
E. Kalnay, “Atmospheric Modeling, Data Assimilation and Predictability”, Cambridge University Press, 2003.
S. Haykin, “Neural Networks: A Comprehensive Foundation”, Pearson Education, 2002.
S. Haykin, “Adaptive Filter Theory”, Mcmillan, 1994.
E.N. Lorenz, Deterministic nonperiodic flow, Journal of the Atmospheric Physics, 20 (1963), 130–141.
A.G. Nowosad, “Novas Abordagens para Assimilação de Dados Meteorológicos”, Tese de Doutorado, CAP-INPE, 2001.
S. Robinson, Toward an optimal algorithm for matrix multiplication, SIAM News, 38, No. 9 (2005).
J. Stoer, R. Bulirsch, “Introduction to Numerical Analysis”, Springer-Verlag, 3rd Edition, 2000.
V. Strassen, Gaussian elimination is not optimal, Numerische Mathematik, 13 (1969), 354–356.
O. Talagrand, Assimilation of observations, an introduction, J. Meteor. Soc. Japan, 75 (1997), 91–209.
R. Todling, S.E. Cohn, Suboptimal schemes for atmospheric data assimilation based on the Kalman filter, Mon. Wear. Rev., 122 (1994), 2530–2557.
Downloads
Published
How to Cite
Issue
Section
License
Authors who publish in this journal agree to the following terms:
Authors retain copyright and grant the journal the right of first publication, with the work simultaneously licensed under the Creative Commons Attribution License that allows the sharing of the work with acknowledgment of authorship and initial publication in this journal.
Authors are authorized to assume additional contracts separately, for non-exclusive distribution of the version of the work published in this journal (eg, publish in an institutional repository or as a book chapter), with acknowledgment of authorship and initial publication in this journal.
Authors are allowed and encouraged to publish and distribute their work online (eg, in institutional repositories or on their personal page) at any point before or during the editorial process, as this can generate productive changes as well as increase impact and the citation of the published work (See The effect of open access).
This is an open access journal which means that all content is freely available without charge to the user or his/her institution. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles, or use them for any other lawful purpose, without asking prior permission from the publisher or the
author. This is in accordance with the BOAI definition of open access
Intellectual Property
All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License under attribution BY.