Oscillation Parameter Estimation via State-Space Modeling of Synchrophasors

This paper presents a technique for the estimation of oscillation parameters via state-space modeling of synchrophasor data. Due to the spectral leakage of synchrophasors estimated by a Fourier transform-based algorithm, the oscillation parameters such as magnitude and frequency will be inherently distorted. Therefore, a state-space model of the instantaneous waveform signal under power system oscillation is derived as an exponentially damped sinusoidal (EDS) signal in order to account for the spectral leakage. The oscillation frequency and magnitude are estimated in real-time by using an unscented Kalman filter (UKF) based on the state-space model. The estimation accuracy performance is validated using the simulation data of sub-synchronous oscillation (SSO). In addition, the efficacy of the proposed method's performance is verified by the application of the proposed method to real-world oscillation events in a wind farm and comparison with the time-frequency analysis.