Locating Forced Oscillation Source Using Granger Causality Analysis and Delay Estimation

Forced power oscillation inflicts great harm on power systems. A fundamental solution is to locate the oscillation source and remove it. This paper proposes a data-driven method to locate oscillation source using the spectral Granger causality analysis and arrival delay estimation of forced oscillation. The application of the spectral Granger causality analysis to locate oscillation source is proposed. The construction of the AR models and the feature selection for causality analysis are given. To provide comparison and confirmation, a method of arrival delay estimation of forced oscillation is proposed to determine the arrival sequence of forced oscillation at different generators. Cross correlation function is adopted to estimate time delays of oscillation arrivals and to distinguish cause generator from effect generators. The proposed data-driven method combines the spectral Granger causality and arrival delay estimation and is validated by several numerical cases. It is shown that the proposed method is model-free and needs only few recorded data.