The use of multilayer perceptron artificial neural networks for the classification of ethanol samples by commercialization region/Utilizacao de Redes Neurais Artificiais Perceptron de Multiplas Camadas para classificacao de amostras de etanol com a regiao de comercializacao

Samples of automotive ethanol, marketed in the northern and eastern regions of the state of Parana, Brazil, underwent physical and chemical tests. Rates were assessed by Multilayer Perceptron (MLP) neural network for classification. For network training, two hundred epochs, a 0.05 learning rate and...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Acta scientiarum. Technology 2016-04, Vol.38 (2)
Hauptverfasser: Romagnoli, Erica Signori, Silva, Livia Ramazzoti Chanan, Angilelli, Karina Gomes, Ferreira, Bruna Aparecida Denobi, Walkoff, Aline Regina, Borsato, Dionisio
Format: Artikel
Sprache:spa
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Samples of automotive ethanol, marketed in the northern and eastern regions of the state of Parana, Brazil, underwent physical and chemical tests. Rates were assessed by Multilayer Perceptron (MLP) neural network for classification. For network training, two hundred epochs, a 0.05 learning rate and a random subdivision of samples in three groups with 70 for training, 15 for test and 15% for validation were employed. Sixty networks were trained from three different initializations. Three networks, one at each start-up, were highlighted and the one with the best performance presented 8 neurons in the hidden layer, with 95 accuracy training, 96 in the test and 96% in validation. The most important variables in classifications, identified by the network, occurred in the following order: alcohol content, density, pH and electrical conductivity. Application of MLP segmented ethanol samples and identified the commercialization regions. Keywords: biofuel, backpropagation, hidden layer, training. Amostras de alcool etilico automotivo, comercializadas nas regioes Norte e Leste do Estado do Parana, foram submetidas a ensaios fisico-quimicos e os valores tabulados foram apresentados a rede neural do tipo Perceptron de multiplas camadas (MLP) para classificacao. Para o treinamento das redes foram utilizadas 200 epocas, uma taxa de aprendizagem de 0.05 e uma subdivisao das amostras, de forma aleatoria, em tres grupos sendo 70 para treinamento, 15 para teste e 15% para validacao. Foram treinadas 60 redes obtidas a partir de tres inicializacoes diferentes. Tres redes (uma de cada inicializacao) foram destacadas, e aquela com melhor desempenho apresentou oito neuronios na camada oculta com 95 de acerto no treinamento, 96 no teste e 96% na validacao. As variaveis mais importantes na classificacao, identificadas pela rede, seguiram a seguinte ordem: teor alcoolico, pH, massa especifica e condutividade eletrica. A aplicacao da rede MLP permitiu a segmentacao das amostras de alcool e, com isso, identificar as regioes de comercializacao. Palavras-chave:biocombustivel, retropropagacao, camada oculta, treinamento.
ISSN:1806-2563
DOI:10.4025/actascitechnol.v28i2.27597