Deep Learning for Remote Monitoring of Power System

The paper comprises two distinct case studies within power systems. The first case study delves into the detection of a sudden load change utilizing an FDR sensor. The second case study addresses the detection of three distinct types of faults in a 9- bus power system. In both case studies, feedforward neural networks are trained for the detection of faults and load changes in power systems. Furthermore, we assess the robustness of these neural networks through simulations and numerical validation, utilizing datasets that are either corrupted by noise or exhibit variations in the initial state probability distribution.