Topology error identification for the NEPTUNE power system using an artificial neural network

摘要

The goal of the North Eastern Pacific Time-Series Undersea Networked Experiment (NEPTUNE) is to construct a cabled observatory on the floor of the Pacific Ocean, encompassing the Juan de Fuca Tectonic Plate. The power system associated with the proposed observatory is unlike conventional terrestrial power systems in many ways due to the unique operating conditions of cabled observatories. The unique operating conditions of the system require hardware and software applications that are not found in terrestrial power systems. This paper builds upon earlier work and describes a method for topology error identification in the NEPTUNE system that utilizes an artificial neural network (ANN) to determine single contingency topology errors.