Fault Location In Electrical Power Distribution Network Using Support Vector Regression and Wavelet Packet Transform Method

This paper presents a fault location protection scheme in a power distribution system. The paper amplifies the importance of condition monitoring in power systems for reliability enhancement of the network. In this paper, a hybrid protection fault locating scheme based on wavelet packet transform (WPT) and support vector regression (SVR) is proposed. The proposed scheme uses the WPT to extract statistical features of the fault and subsequently, the SVR technique is employed to estimate the location of the fault. Furthermore, the proposed WPT-SVR scheme is tested using 1/2 cycles for signal analysis. An Eskom 132 kV power system line is modelled using the Digsilent platform. Thereafter, various types of fault cases are investigated from the modelled network. The fault classification scheme is tested using a machine learning platform WEKA. The results obtained show that the different impedance variations do not affect the efficiency of the scheme.