Phasor Data Compression with Principal Components Analysis in Polar Coordinates for Subsynchronous Oscillations

In the previous studies, the phasor data are processed as the amplitudes and the phases separately since existing methods are in the field of real numbers. To take advantage of the correlations between amplitudes and phases of phasors, a phasor principal components analysis (PPCA) method in polar coordinates is proposed in this paper. In PPCA, the conventional principal components analysis (PCA) is improved and extended to the field of complex numbers to process the phasors in polar coordinates. The actual PMU data measured in a subsynchronous oscillation incident are used to verify the compression performance of PPCA which is also compared with PCA. The result demonstrates that PPCA can achieve better performance by making use of the correlations between amplitudes and phases of phasors.