A wavelet packet transform approach for locating faults in distribution system

Accurate location of faults in a distribution system is a need of present scenario in order to provide uninterrupted supply to the customers. Research is under way to increase the accuracy of the fault in distribution system. It should also be kept in mind that transmission line fault location algorithms cannot be directly employed on distribution system because of the configuration of distribution system. In the present work, the authors have developed an algorithm for locating all ten types of faults by extracting the features of voltage signals by wavelet packet transform and subjecting it to artificial neural network. The voltage samples have been collected after extensive simulation of a sample distribution system over different range of fault resistance, inception angle in PSCAD (Power System Computer Aided Design) software. The results thus obtained from the proposed algorithm are found to be very promising. The algorithm has been developed keeping in view the pragmatics hurdle of the multiple estimation which have been successfully tackled.