Classificação de Distúrbios de Qualidade de Energia Elétrica Empregando Redes Neurais Artificiais

Marcel Ayres de Araujo¹; Rogério Andrade Flauzino¹; Oureste Elias Batista¹; Lucas Assis de Moraes¹; Cesar Henrique Rossinoli Martins¹

¹ Universidade de São Paulo (USP)

doi:10.20906/CPS/SBSE2016-0427

Resumo

Increasingly are necessary studies and methods for analysis and classification of disturbance related to Power Quality (PQ) for the electric utilities are able to identify and control the sources of such disorders in order to eliminate them or at least mitigate them as efficiently as possible and thus provide a better product to consumers, whether residential, industrial or commercial. Thus, this study presents a classification method for disturbance related the PQ with the use of an Artificial Neural Network (ANN), which will be able to classify five types of disturbance, that are: voltage sags, voltage swell, impulsive transients, momentary interruption, and harmonic distortion.

Palavras-chave: artificial neural network; multilayer perceptron; power quality; power system