Modal Identification Of Power System Low Frequency Oscillation Based On Improved VMD-HT

In recent years,with the scale of interconnected power system is more and more large,the power grid construction in our country into a long distance,large capacity,ac/dc hybrid system,system operation is complicated,the low frequency oscillation phenomenon occurs frequently,poses a great threat to the safe operation of the power system.At the same time,the wide application of WAMS system makes the modal identification of low-frequency oscillation based on measurement signal get more attention.The purpose of this paper is to combine VMD decomposition and HT transformation to realize low-frequency oscillation modal identification.Firstly,analysis the identification parameters of the Hilbert spectrum with the oscillation parameters,as HT connot identify multi oscillation mode,consider using mode decomposition method to develop this defect.Use two kinds of oscillation mode decomposition algorithm: EMD and VMD to identify the simulation signal of different parameter combinations,the VMD algorithm has better applicability on the lowfrequency oscillation modal identification.A new method for low frequency oscillation modal identification based on VMD-HT is proposed.Then,improve the shortcomings of VMD-HT: endpoint effect,mode mixing and false component phenomenon caused by inaccurate modal number,The endpoints of the oscillating signal are treated BP neural network and mirror extension,the endpoint effect of VMD-HT method is inhibited.According to the characteristics of low-frequency oscillation signal,a strategy to determine the number of oscillation modes based on the entropy of instantaneous-amplitude is proposed.Finally,Use improved VMD-HT method to identify the parameters of multi-machine simulation signal and forced-oscillation measured signal,the results show that the proposed method based on improved VMD-HT is effective and accurate in identifying low-frequency oscillation mode parameters.Improved VMD-HT method has better noise resistance than Prony method,and can describe the non-stationary characteristics of oscillation modal parameters accurately,has certain engineering application value.