Wavelet-ANN based classification of HVDC converter faults

High voltage direct current transmission (HVDC) system is preferred for bulk power transmission over a long distance. HVDC system consists of two converters on either ends of the system which are acting as a rectifier and an inverter. These converters are sensitive and are exposed to various internal faults such as back fire, arc through, misfire and commutation failures. In this paper, a fast technique is employed to detect these internal faults by employing wavelet analysis combined with artificial neural network. Wavelets are used to extract the transients in fault currents and artificial neural network is employed for automatic classification of faults. Performance of the classifier is tested for various internal of the converter and is found satisfactory.