A wavelet packet transform approach for locating faults in distribution system

Wavelet‐based extreme learning machine for distribution grid fault location

“…da Silva et al [14] estimated the location of high-impedance faults on a Brazilian distribution grid where three-phase currents along with the energies of coefficients obtained from discrete wavelet transform (DWT) were employed to train and test the ANNs. Another technique located different types of faults by extracting the features of voltage signals employing wavelet packet decomposition and fetching those to ANN [15]. A hybrid technique located faults employing ANN/SVM while DWT extracted the features of one cycle post-fault voltage data and genetic algorithm (GA) selected the best features to reduce computation burden [16].…”