Artificial neural network based fault identification of HVDC converter

This paper investigates the possibility of using a neural network for HVDC converter protection. Based on the ability of this network to distinguish reliably between different types of faults that may occur in a converter, the feature can be suitably integrated with an ANN based controller to improve the dynamic response of an AC-DC power system. In this paper, three new neural network based methods to distinguish different faults in an HVDC converter are proposed and comparison between them is made under different system perturbations and faults. The identifier is tested for HVDC with a strong and weak AC side. The method is independent of the operating mode of the converter. The proposed fault identifier can be used to design an integrated ANN based controller with fault identifier for the HVDC system.