Research On Analysis Method Of Power System Broad-band Oscillations Based On Measurement Signal

For ensuring the safe and stable operation of the power system,the stability of power system is one of basic conditions.At present,the power system is facing many challenges such as the global energy Internet,the stability of the AC-DC hybrid power grid,and the access to large-scale renewable energy.This also makes the oscillation problem in the power system increasingly prominent.Low-frequency oscillation,subsynchronous oscillation and super-synchronous oscillation of power system were discovered early.Their mechanisms have been studied thoroughly and there are many types of suppression methods.Some domestic scholars took the lead in proposing the concept of Broad-Band Oscillations analysis for the above three oscillation problems,but no direct definition was given.In recent years,a new type of oscillation problem—ultra-low frequency oscillation—has appeared in power systems where hydropower accounts for a large amount,which has caused research by domestic and foreign scholars.The detection and analysis of oscillation modal parameters is an important part of formulating suppression strategies and online monitoring of oscillations.At present,there are many researches on the analysis methods of various oscillations separately,and the analysis methods of various oscillations have certain similarities.Further research on the analysis method of Broad-Band Oscillations is of great significance to the detection and analysis of various oscillations in power systems and online monitoring.In this paper,the ultra-low frequency oscillation problem is firstly summarized in the analysis of power system Broad-Band Oscillations,and the definition of power system Broad-Band Oscillations analysis is put forward in summary.It is pointed out that its purpose is to unify the detection,analysis and monitoring methods of the above four types of oscillation,so that it is more efficient when developing software for detection,analysis and monitoring.At present,the wide area measurement system(WAMS)is becoming more and more complete,which provides a reliable means for monitoring,analysis and control of large-scale interconnected power grid systems,and also provides a guarantee for the engineering application of Broad-Band Oscillations analysis.Then,the pre-processing of the Broad-Band Oscillations signal is studied.Based on the Kaiser window function,a multi-band linear phase filter suitable for the analysis of pan-band oscillations is designed.An ideal Broad-Band Oscillations signal is constructed,its performance is verified,and four types of oscillation signals are separated.By comparing and analyzing the causes and suppression methods of the end effect problem of the variational mode decomposition algorithm and the empirical mode decomposition algorithm,the extension method for suppressing the variational mode decomposition problem in this paper is determined.And then,according to the physical properties of the Broad-Band Oscillations signal,the selection of parameters and the deficiencies of the algorithm when applying the variational modal decomposition algorithm to the analysis of the power system’s wide band oscillation are studied.Based on the singular value decomposition method,the adaptability of the variational modal decomposition algorithm is improved,and the modal identification link is improved by the Prony algorithm.A SVD-VMD identification algorithm for pan-band oscillation analysis is proposed.By establishing an ideal signal model,the effectiveness and superiority of the algorithm are verified by comparison.Finally,the oscillating events occurring in two actual power systems are studied,and their modal parameter changes from the start-up stage to the stable oscillation stage are analyzed,which further verifies the effectiveness of the proposed algorithm in practical applications.