Você está no menu de acessibilidade

Utilize este identificador para citar ou criar um link para este item: http://acervodigital.unesp.br/handle/11449/9730
Título: 
Electrical load forecasting formulation by a fast neural network
Autor(es): 
Instituição: 
Universidade Estadual Paulista (UNESP)
ISSN: 
0969-1170
Resumo: 
The objective of this work is to develop a methodology for electric load forecasting based on a neural network. Here, backpropagation algorithm is used with an adaptive process that based on fuzzy logic and using a decaying exponential function to avoid instability in the convergence process. This methodology results in fast training, when compared to the conventional formulation of backpropagation algorithm. The results are presented using data from a Brazilian Electric Company, and shows a very good performance for the proposal objective.
Data de publicação: 
1-Mar-2003
Citação: 
Engineering Intelligent Systems For Electrical Engineering and Communications. Market Harboroug: C R L Publishing Ltd, v. 11, n. 1, p. 51-57, 2003.
Duração: 
51-57
Publicador: 
C R L Publishing Ltd
Palavras-chaves: 
  • load forecasting
  • short term
  • neural networks
  • backpropagation
  • fuzzy logic
Endereço permanente: 
Direitos de acesso: 
Acesso restrito
Tipo: 
outro
Fonte completa:
http://repositorio.unesp.br/handle/11449/9730
Aparece nas coleções:Artigos, TCCs, Teses e Dissertações da Unesp

Não há nenhum arquivo associado com este item.
 

Itens do Acervo digital da UNESP são protegidos por direitos autorais reservados a menos que seja expresso o contrário.