Detection and Classification of Transmission Line Faults Using Combined Features of Stockwell Transform, Hilbert Transform, and Wigner Distribution Function

Abstract

Performance of transmission and distribution power network is affected by faults. Hence, there is a need to develop a smart protection mechanism that not only provides widespread protection to the line but also stands economical and equally reliable with the growing need. Therefore, a quick diagnosis of faults and accuracy in the estimation of fault location not only will help in a faster maintenance but also is required for quick restoration of power supply. The present work mainly focusses on theoretical aspect. From the literature survey it has been found that field-related issues are addressed rarely. In the field conditions, while supervising supply system, it was observed that the system trips without any actual fault such as switching-on of capacitor bank, charging of power transformer, etc. Hence, there is a need to develop a protection scheme that not only trips the feeder/power line on actual fault condition but also takes less reaction time during the occurrence of a fault. Therefore, in order to have a fast and efficient diagnosis of a fault, an algorithm has been developed to program a relay that overcomes the demerits of existing relays using signal processing techniques, such as the Stockwell Transform, Hilbert Transform, and Wigner Distribution Function. A test transmission line representing a large area utility network has been used to test faults like symmetrical and unsymmetrical. Also, switching events, variation in fault impedance and incidence angle has been analyzed. The proposed algorithm has been implemented under MATLAB/Simulink environment. The simulation results show a good agreement with theoretical concepts for classification and detection of faults under different conditions, such as change in line length, reverse power flow condition, and switching transients.