Detecção e classificação de faltas em sistemas de distribuição de energia elétrica usando a análise multirresolução e a rede neural artificial ARTMAP-fuzzy

Carregando...
Imagem de Miniatura

Data

2019-07-05

Autores

Bernardes, Haislan Ranelli Santana

Título da Revista

ISSN da Revista

Título de Volume

Editor

Universidade Estadual Paulista (Unesp)

Resumo

Esta pesquisa apresenta o desenvolvimento de uma ferramenta para a detecção e classificação de curtos-circuitos em sistemas de distribuição de energia elétrica, a qual é baseada no uso combinado da análise multirresolução e rede neural ARTMAP-fuzzy. A análise multirresolução permite a identificação de singularidades nas oscilografias e a rede da família ART garante ao sistema classificador a capacidade de aprendizado contínuo de novos padrões sem perder o conhecimento previamente adquirido. Todo o processo de diagnóstico é realizado em uma única etapa, reduzindo o custo computacional da metodologia. A eficiência do sistema é verificada por uma análise direta, na qual se contabiliza a quantidade total de acertos, e por uma avaliação comparativa, a qual envolve a substituição da rede ARTMAP-fuzzy pela rede ARTMAP-Euclideana. Resultados mostram que o sistema é eficiente, sendo capaz de detectar e classificar 100% das falhas elétricas.
This research presents the development of a tool for the detection and classification of short circuits in electric power distribution systems, which is based on the combined use of multi-resolution analysis and fuzzy ARTMAP neural network. The multiresolution analysis allows the identification of singularities in the oscillographs and the ART family network guarantees to the classifier the ability to continuously learn new patterns without losing the previously acquired knowledge. The entire diagnostic process is performed in a single step, reducing the computational effort. The efficiency of the system is verified by a direct analysis, in which the total number of hits is counted, and by a comparative evaluation, which involves the replacement of the fuzzy ARTMAP network by the Euclidean ARTMAP network. Results show that the system is efficient, being able to detect and classify 100% of the electrical faults.

Descrição

Palavras-chave

Inteligência artificial, Redes neurais artificiais, Detecção de faltas de curto-circuito, Classificação de curto-circuito, ARTMAP-fuzzy, Artificial intelligence, Artificial neural networks, Detection of short circuit faults, Classification of short circuit, Fuzzy ARTMAP

Como citar