Spectral-based dynamic analysis for estimation of mode shapes

This paper presents spectral-based dynamic approach for mode shape estimation by using time-domain signals. The approach in this paper is applied on software simulated signals which partially resembles the PMU signals. Power System Toolbox (PST) is used to simulate IEEE 16 machine 68 bus test system to generate multiple signals. Empirical mode decomposition (EMD) decompose multimode signal into separate modes but perfect decomposition by EMD is not always guaranteed therefore two mode decomposition techniques are used at two stages for effective mode decomposition. Power Spectral Density (PSD) and Cross Power Spectral Density (CPSD) tools are used to process decomposed modes for estimation of mode frequency and mode shape respectively. The mode shape results are compared and verified with the conventional approach i. e., eigenvalue analysis. Comparison shows that the approach is effective in mode shape estimation and hence it could also be applied on real-time PMU signals.