A new approach for online identification of low frequency oscillation modes of power system based on the improved Hilbert-Huang transform

A new nonlinearity, instability and non-stationary signal processing method named improved Hilbert-Huang transform was proposed to analyze the measured oscillatory signal from wide area measurement system. Based on this method, the mode mixing of the measured signal in decomposition process was removed and the scheme has advantages of good performance of anti-interference. The measured signal was decomposed into a series of models and identified the amplitude, frequency and damping of low-frequency oscillation signal. The measured signals from simulation model and actual power system indicate the validity and feasibility of the proposed method.