An Evaluation of Algorithms for Synchrophasor Missing Data Recovery

Abstract

Rising numbers of Phasor Measurement Units (PMUs) are being installed in the North American power grid, allowing for a large amount of system dynamics to be collected at high sampling rates. The data collection process requires a phasor data network, which due to congestion, results in missing data affecting the reliability of applications using PMU data. Low-rank matrix methods have been proposed as tools to recover missing PMU data. The PMU data recovery performance of three existing algorithms, namely, the Singular Value Thresholding (SVT) algorithm, the OnLine Algorithm for PMU data processing (OLAP), and the Jones-Pal-Thorp extrapolation algorithm, as well as a modified version of the OLAP algorithm proposed in this paper, is investigated using historic PMU data from the New York and New England power systems.