Título: Identificação de Cargas Domésticas Utilizando uma Rede ART a Partir de Amostras de Potência Ativa e Reativa
Autores: Silva, Pedro Bittencourt e; Seixas, José Manoel de; Calôba, Luiz Pereira
Resumo: This work trusts in known methodologies to feature extraction from on-off events, of residential appliances, from active and reactive power data series, it’s proposed an ART network clustering to identify those loads based on the related event variation of active and reactive power. It’s shown that unsupervised learning techniques could be used for appliance load identification, in special ART theory due its plasticity and ease to work on dynamic environments.
Palavras-chave: NIALM; Unsupervised Training; Artificial Neural Networks; Smart Grid
Código DOI: 10.21528/CBIC2013-254
Artigo em pdf: bricsccicbic2013_submission_254.pdf
Arquivo BibTex: bricsccicbic2013_submission_254.bib