Artificial neural networks, Fault location, Distributed generation, Medium voltage distribution system
Abstract
This work aims to address the problem of fault location in a medium voltage distribution system in the presence of Distributed Generation (DG), applying artificial intelligence. The chosen method is learning-based through Artificial Neural Networks (ANNs). The methodology consisted of simulating three-phase faults via PSCAD/EMTDC software, pre-processing (normalization and formatting of training and test sets), empirical tests, routine tests without DG, routine tests with DG, and choosing the best ANN topology. The results presented illustrate the applied tests and the best ANN topology found.
