HVDC Converter Fault Discrimination using Probabilistic RBF Neural Network Based on Wavelet Transform

The progressive development in HVDC transmission systems enhances the need of implementing efficient protection schemes to distinguish the minimum faulty part of the system and to relief the stressed equipment. In this paper, at first different types of converter faults are introduced, and characteristics of each fault are investigated. These faults have been distinguished using of wavelet transform and probabilistic RBF neural network. Based on the proposed algorithm, a high speed protective control decision with small computational time could be performed in 12.5ms. Simulation of CIGRE standard HVDC system proves the proposed technique effectiveness and