Determination of mode shapes in PMU signals using two‐stage mode decomposition and spectral analysis

Abstract

This study presents a dynamic approach for determining mode shape using time-domain signal. Signal processing techniques, mode decomposition, and spectral analysis are used here. The quality of signal affects the performance of spectral analysis, especially for estimation of low-frequency modes. Therefore, before applying spectral analysis, low-frequency modes are extracted using mode decomposition technique, so that the decomposed modes (DMs) may indicate centre frequency in its spectrum. In the study, two-stage mode decomposition approach is proposed for accurate and effective mode decomposition. The power spectral density (PSD) and cross-PSD tools are used to process DMs for estimation of mode frequency and determination of mode shape, respectively. The proposed dynamic approach is tested on simulated signals of IEEE 16-machine 68-bus test system and real-time phasor measurement units (PMUs) signal. The results obtained using proposed dynamic approach on simulated signals are compared with those obtained by steady-state approach, i.e. eigenvalue analysis.