A Time-Frequency Transform based Fault Detection and Classification Methodology for Transmission Lines

The inevitable events in power systems such as faults should be detected and resolved quickly to maintain system reliability. This paper proposes a Time-Frequency transform (wavelet transform) based fault detection and classification methodology using current signals. The Daubechies wavelet has been used to extract the features of the current signals. The proposed method detects a fault using the first level decomposition coefficients using wavelet transform, while the fault is classified by using the maximum values of the detail coefficients and logical analytical techniques. The proposed methodology is validated on a test model developed in the MATLAB Simulink environment. The performance of the proposed methodology has been verified under different fault configurations for different fault locations, resistances and inception times. The algorithm is also validated for a load change at the time of fault inception. The results show that the proposed methodology is accurate and reliable in fault detection and classification and can help in taking appropriate decisions to enhance the reliability of power system.