Dynamic Phasor and Frequency Measurements by an Improved Taylor Weighted Least Squares Algorithm

Abstract

One of the most accurate phasor estimation procedures recently proposed in the literature is the so-called Taylor weighted least squares (TWLS) algorithm, which relies on a dynamic phasor model of an electrical waveform at nominal frequency. In this paper, an extension of the TWLS algorithm [called generalized TWLS (GTWLS) algorithm] to a generic (not only nominal) reference frequency is described and the accuracies of the returned estimates are analyzed through meaningful simulations, performed in different steady-state and dynamic testing conditions according to the IEEE Standard C37.118.1-2011 about synchrophasor measurement for power systems and its Amendment IEEE Standard C37.118.1a-2014. It is shown that the accuracy of the total vector error (TVE), frequency error (FE), and rate of change of frequency error (RFE) normally decreases as the deviation between the reference frequency and the true waveform frequency decreases. Furthermore, a two-step procedure for accurate estimation of the phasor parameters is proposed. In the first step, the waveform frequency is estimated by a classical interpolated discrete Fourier transform (IpDFT) algorithm. The second step then returns an estimate of the phasor parameters by applying the TWLS algorithm based on the frequency estimate returned by the first step. It is shown that the proposed procedure, called the GTWLS–IpDFT algorithm, can comply with the $P$ -class or the $M$ -class of performances in all the considered testing conditions when an appropriate number of waveform cycles is considered and the most significant disturbances are removed from the analyzed waveform. Finally, uncertainties of the proposed estimators and the $\mathrm{IpD}^{2}$ FT algorithm recently presented in the literature are also compared.