Fault Location after Fault Classification in Transmission Line Using Voltage Amplitudes and Support Vector Machine

In this paper, a fault diagnosis scheme is proposed in the transmission line, by which the faults can be classified and located correctly and rapidly using voltage amplitudes and support vector machine. A high voltage power system transmission line is simulated by MATLAB to produce a fault data set. The three-phase fault voltages are obtained at one terminal point of the line. The amplitudes of the three-phase fault voltages will be applied as the fault features to train the support vector classification (SVC) and realize fault classification after the three-phase fault voltages pass through a low-pass filter to remove the noise. After knowing the fault type, the voltage amplitude of the phase where the fault occurs will be used as the fault feature to train the support vector regression (SVR) and realize the fault location. Compared with other fault classification and fault location schemes, the proposed fault diagnosis scheme needs less information to provide one hundred percent classification accuracy and high accurate estimations of fault location. Simulations and comparisons show the proposed scheme is better than other schemes.