Non-Pilot Protection of the Inverter-Dominated Microgrid using Artificial Neural Networks

Abstract

Reliable protection of the inverter-dominated microgrid is a technical challenge considering the peculiar behavior of the power electronics converters interfacing distributed energy resources with the grid. Traditional protection strategies fail to reliably detect the direction and type of faults in such microgrids. This paper introduces a non-pilot protection strategy using artificial neural networks. The proposed method is fast, secure, and robust against the microgrid mode of operation and parameters. Comprehensive simulation studies are conducted in PSCAD/EMTDC software to investigate the performance of the proposed protection strategy under different fault scenarios. The results indicate that the proposed protection strategy achieves 100% accuracy in fault direction detection and fault type classification.