An algorithm for the protection of distribution feeders using the Stockwell and Hilbert transforms supported features

Abstract

Faults' recognition in the distribution feeders (DFs) is extremely important for improving the reliability of the distribution system. Therefore, this paper proposes a technique to identify the faults on the DF using the Stockwell Transform (ST) dependent variance feature and Hilbert transform (HT) by utilizing current signals. By element to element multiplication of the H-index, we compute using HT aided decompositions of current waveforms and VS-index, and calculate through ST aided decomposition of current waveforms. By utilizing the decision rules, various faults are classified. Different faults studied in this work are line to ground, double line, double line to ground and 3-Φ to ground. For high fault impedance, this technique is effectively utilized. Furthermore, variations in the fault incidence angles are also utilized to test the performance of the proposed technique. To perform the proposed algorithm, a IEEE-13 bus system is developed in MATLAB/Simulink software. The algorithm effectively classified the faults with accuracy greater than 98%. The algorithm is also successfully validated on the IEEE-34 bus test system. Furthermore, the algorithm was successfully validated on the practical power system network. It is recognized that the developed method performed better than the discrete Wavelet transform (DWT) and ruled decision tree based protection scheme reported in various literature.