Distance protection using a time-frequency filtering technique

The paper presents a new time-frequency formulation similar to the conventional discrete S-transform but with a significantly increased speed of computation suitable for digital protection of power transmission lines. The new formulation uses frequency scaling, band pass filtering, and interpolation techniques to achieve fast and accurate estimation of fault location in an interconnected power system network. The new transform uses half cycle of the data prior to the initiation of the fault and half cycle data after the fault to yield impedance measurement at the relay location with significant accuracy even in the presence of decaying DC components, harmonics, and distortions due to noise. Several fault case studies have been performed to validate the performance of this new algorithm for different fault resistances and fault location along the protected line. For comparison, a recently proposed robust Gauss-Newton algorithm is used to produce impedance trajectories for the case studies attempted in this paper instead of the widely used discrete Fourier transform, which gives erroneous results in the presence of decaying DC and harmonic components. Copyright (c) 2014 John Wiley & Sons, Ltd.