Identifyng Low Frequency Oscillation Modes In Power System Using Multivariate Empirical Mode Decomposition

The low frequency oscillation phenomenon in power system has always been one of the main factors threatening the security operation of power grid,if the oscillation does not subside spontaneously and can not be suppressed rapidly,the continuous oscillation would cause the power grid to collapse and a series of security accidents.Therefore,the rapid and accurate identification of low frequency oscillation in power system can help power operators to understand the cause of low frequency oscillation and take corresponding inhibition measures and maintain the security operation of power grid.With the large-scale configuration of the synchronous Phasor Measurement Unit in the power grid,the identification methods based on wide area measurement information have made great strides in the field of low frequency oscillation analysis,but,most of these methods are limited to identifying single channel measurement signal,thus the identification results are susceptible to the influence of the observability of oscillation modes and measurement noise,and the oscillation mode between different signal channels is difficult to calibrate.Aiming at this shortcoming,this paper adopts a power system dominant oscillation pattern identification method using multivariate empirical mode decomposition(MEMD).By using multivariate empirical mode decomposition,the synchronous decomposition of multi-channel measurement signal is realized,and the intrinsic mode functions(IMFs)corresponding to different oscillation modes are obtained,and then the relative energy value of each IMF is calculated by using the Teager energy operator.Based on the relative energy weight,the IMFs corresponding to the dominant oscillation mode concerned by the low frequency oscillation is selected,then the prediction error method(PEM)is adopted because of the advantage of the small prediction error,based on the algorithm principle of PEM,the current method for calculating the oscillation parameters of the measured data has improved,so as the oscillation frequency and damping ratio of the dominant oscillation mode are solved accurately.Finally,through the simulation data of IEEE 68 nodes test system and the measured data of PMU wide area of Liaoning power grid,the method proposed in this paper is verified,and the analysis results verified the accuracy and validity of the method proposed.