Analyzing Power Oscillating Signals With the O-Splines of the Discrete Taylor–Fourier Transform

Abstract

Power oscillating signals are analyzed with the Discrete Taylor-Fourier Transform (DTFT). This is implemented by modulating the impulse responses of its low-pass differentiators, defined by a family of O-splines. This implementation reduces its computational complexity since in practice only a small subset of filters is applied. The estimated parameters provide richer dynamic information than the traditional methods, in particular, a space representation for each dynamic component, and detection of frequency modulated events. Their estimation performance is assessed through the new proposed Total Phasor Error. To illustrate its application, and its progressive accuracy, this technique is applied to observe the synchrophasor estimates of voltages and to separate the electromechanical modes of an oscillation in a real power system. In conclusion, this multiresolution technique provides a series of increasingly precise solutions for time-frequency separation of oscillations with fluctuating frequency.