Metodologia para detecção e localização de faltas em rede real de distribuição considerando a inserção de veículos elétricos

Abstract

The present master’s thesis proposes an application of artificial neural networks in the problem of detecting and locating high impedance defects in a real electrical energy distribution system with the presence of electric vehicles. For this, electric vehicles with their charging stations were used and the daily charging curve of electric vehicles was varied in order to assess the impact on the problems of detecting and locating high impedance defects, on the premise of adaptive protection through application of RNA. Among the curves in the specialized literature, three were chosen, with special emphasis on the type of recharge in which the electric vehicle is considered, both in the charging condition and in the condition of power injection in the distribution network. In addition to the charging curves, in order to increase the applicability of the method, the load levels were also varied for the simulation of power flow via OpenDSS. Finally, we investigated different types of topologies for the installation of smart meters installed throughout the electrical energy distribution system studied and the results were analyzed using evaluation metrics and descriptive statistics tools. The case study considers a real electrical network from a Brazilian distribution concessionaire.