Busca avançada
Ano de início
Entree
(Referência obtida automaticamente do Web of Science, por meio da informação sobre o financiamento pela FAPESP e o número do processo correspondente, incluída na publicação pelos autores.)

Islanding detection of synchronous distributed generators using data mining complex correlations

Texto completo
Autor(es):
Gomes, Eduardo A. P. [1] ; Vieira, Jose C. M. [1] ; Coury, Denis V. [1] ; Delbem, Alexandre C. B. [2]
Número total de Autores: 4
Afiliação do(s) autor(es):
[1] Univ Sao Paulo, Sao Carlos Sch Engn, Dept Elect & Comp Engn, Ave Trabalhador Sao Carlense 400, BR-13566590 Sao Carlos, SP - Brazil
[2] Univ Sao Paulo, Inst Math & Comp Sci, Ave Trabalhador Sao Carlense 400, BR-13566590 Sao Carlos, SP - Brazil
Número total de Afiliações: 2
Tipo de documento: Artigo Científico
Fonte: IET GENERATION TRANSMISSION & DISTRIBUTION; v. 12, n. 17, p. 3935-3943, SEP 30 2018.
Citações Web of Science: 2
Resumo

This study proposes a novel method based on data mining for islanding detection of synchronous distributed generators. The method uses a new technique called data mining of code repositories (DAMICORE), which is a powerful data mining tool for detecting patterns and similarities in various kinds of datasets. In addition, a trip logic was developed in this proposal in order to detect islanding and disconnect the distributed generator. One of the most relevant features of the proposed method is its capability to generalise, which reduces the need of big datasets for training purposes. This approach has been tested with islanding, load switching and fault simulations, presenting promising results concerning performance, as well as detection time. General results showed a better performance of the method if compared to traditional anti-islanding protection schemes, such as frequency-based relays. (AU)

Processo FAPESP: 17/16742-7 - Microrredes: identificação e solução de problemas em sistemas híbridos de geração distribuída
Beneficiário:Denis Vinicius Coury
Modalidade de apoio: Auxílio à Pesquisa - Regular