The Effects of Forced Oscillation Frequency Estimation Error on the LS-ARMA+S Mode Meter

The least squares autoregressive moving average plus sinusoids (ls-arma+s) mode meter has been previously shown to successfully estimate power system electromechanical modes without bias in the presence of a forced oscillation (FO). In order to do this, it requires knowledge of the FO frequency, which is estimated as a byproduct of the periodogram-based FO detection scheme. Previous work has not fully addressed the effect that FO frequency estimation error has on the effectiveness of LS-ARMA+S. This letter demonstrates that if a FO is powerful enough, its frequency must be estimated with immense accuracy, something that the existing technique may not be able to achieve. Here it is demonstrated that easy-to-implement alternative frequency estimation algorithms can overcome this, resulting in the recommendation that for any future use of the LS-ARMA+S mode meter, the FO frequency be estimated with an algorithm separate from that which is used for FO detection.