Application of artificial neural networks for the detection and localization of faults in electrical distribution systems: A case study in the canton of La Joya de los Sachas

Electrical distribution systems play an important role in delivering power to consumers, and their reliability is of utmost importance. This paper presents a comprehensive study on the application of artificial neural networks (ANNs) for fault detection and localization in electrical distribution systems, focusing on a case study in the canton of La Joya de los Sachas, Ecuador. The research employs a multilayer perceptron (MLP) neural network architecture to analyze voltage and current measurements from various nodes within the distribution network. The ANN model is trained using a dataset of simulated fault scenarios, encompassing different fault types and locations. Results demonstrate that the ANN-based method achieves high accuracy in both fault detection and localization, with an overall success rate of 98.5% for fault detection and a mean localization error of less than 50 meters. The study also compares the ANN approach with traditional fault detection methods, highlighting its superior performance in terms of speed and accuracy. Furthermore, the paper discusses the practical implications of implementing this technology in the La Joya de los Sachas distribution network, including potential improvements in system reliability and reduction in outage duration. This research contributes to the advancement of smart grid technologies and provides valuable insights for utility companies seeking to enhance their fault management capabilities.