Fuzzy based fault classifier and time domain reflectometry based fault locator for underground cables

Abstract

Cables are insulating wires used for transferring power or transmitting signals. If any fault occurs in the cable then, it causes detrimental effects to power system equipments. The technical difficulty is to find, analyze, and classify faults. In case of underground cables it will be a time consuming and expensive process to repair the faults or replace the cables. This paper aims to analyse the faults by creating the various types of short and open circuit faults, classify them by taking in account of grounding, and to locate the distance at which the fault is occurred. The underground cable is simulated using MATLAB/ Simulink, the faults are created and the Fourier transform of voltage and current are observed at the source end. The generated fault phase current data from the simulation has been normalized using fault index is the input to the Mamdani fuzzy logic controller. The defuzzified output is verified for different phase current values which classify the nature of the fault. Two fuzzy classifiers like ground and phase faults are developed which holds good for different range of cables. A methodology to find the fault location in long transmission line is attempted using PSPICE, by creating short circuit fault. The distance is determined by using velocity of propagation of travelling waves, which is realized using impulse voltage wave generator which is based on impedance.