Fault Classification, Location in a Series Compensated Power Transmission Network using Online Sequential Extreme Learning Machine

This paper presents a unique approach for fault classification and fault location estimation in a series compensated power transmission network. The method uses Online Sequential Extreme Learning Machine (OSELM) neural network in conjunction with determinant function for classification and location of faults. In this technique current, voltage and power phasor values are measured for half cycle duration from fault inception at the relaying end of the faulted transmission line. The determinant function is used to extract fault features. Optimal fault feature set is selected using sequential forward selection (SFS) method. The selected features are fed as input to OSELM for fault classification and fault location estimation. The proposed method is extensively tested in various types of fault scenarios in a three machine nine bus power transmission network with series compensation simulated in DigSilent Power factory. The results prove that the proposed scheme is highly accurate and fast in fault classification and fault location estimation.