Signal processing techniques for synchrophasors considering short-circuit signals: a comparative study

Synchrophasors represent the next step in the level of information available considering operation, protection, control and stability of power systems. Among the various different uses of synchrophasors, currently numerous studies have proposed improved power system protection strategies using phasor measurement units. As the core of synchrophasor construction is the signal processing technique used to estimate the local phasor, it is important to adopt proper estimation algorithms that improve protection algorithm performances. Therefore, this study compares the performance of different signal processing techniques (discrete Fourier transform, TLS-estimation of signal parameters via rotational invariance technique, amplitude modulation model, recursive least-squares, improved recursive Newton-type algorithm and Robust Gauss-Newton) to estimate amplitude, phase and frequency of AC signals, including synthetic and fault signals. Several tests were performed, and characteristics such as precision, tolerance to noise and harmonic components, transient tracking and computational burden were analysed for each technique. The results showed that windowed techniques presented better precision than recursive techniques.