A Balancing Voltage Transformation for Robust Frequency Estimation in Unbalanced Power Systems

Yili Xia,Kai Wang,Wenjiang Pei,Danilo Y. Mandic

DOI: https://doi.org/10.1109/apsipa.2014.7041682

2014-01-01

Abstract:This paper addresses the detection of the fundamental frequency of power systems under unbalanced and distorted conditions. By using the second order information, both the autocorrelation and pseudo-autocorrelation, within the Clarke's transformed voltage, a novel balancing voltage transformation (BVT) is proposed to accurately detect the underlying phase angle evolution of the positive sequence component. This removes the biggest obstacle in current power systems and makes possible to use any frequency estimator for single-tone exponential on unbalanced power systems. The robustness of the proposed phase angle detection technique is illustrated for two well-known and efficient frequency estimators, that is, a discrete Fourier transform (DFT) coefficient interpolation method [1] and the weighted linear predictor (WLP) [2]. A window technique is used to cater for the fast and computationally affordable frequency estimation purposes. Simulations over a range of unbalanced conditions, including voltage dips and swells, frequency deviations and the presence of higher order harmonics support the analysis.