The Monitoring Analysis Research Of Low Frequency Oscillation In Power System Based On HHT Method

As “western electricity sent to east, mutual supply of south and north”,our countyhas been into the era of large power grid. In large power system, the low-frequencypower oscillation occurred in the regional power grid tie lines on frequent, this low-frequency power oscillation will induce large era units swing, even serious paralysis,has threat to the safe and the stable operation of the power system. How effectivelymonitoring the low-frequency oscillation and promptly take reasonable measurementsto suppress the suffering is the current hot research topic in the field of power system.Algorithm based on the modal of traditional low frequency oscillation analysismethod has been unable to satisfy the demands of very large scale power systemstability analysis, using the wide-era measurement system(Wide-Area MeasurementSystem, WAMS) monitoring analysis of the measured signal to low-frequencyoscillation has become a trend. For the measured signal oscillation around the trackfairly reflect the dynamic characteristics of the system, contains the eigen modalinformation and excitation mode characteristic information, is a kind of typical non-stationary random signal; Therefor, this paper use a non-stationary signal processingalgorithm--HHT(Hilbert Huang Transform, HHT) to monitoring and analysis of lowfrequency oscillation, from the viewpoint of engineering practice and surrounding thealgorithm and the speed of the online analytic research on the following three aspects:1.Research the improved Hilbert Huang Transform algorithm, solved the EMDdecomposition of endpoint effect and modal aliasing problems, the two problems relatedto HHT algorithm of online identification problem of the accuracy of modal parameters,especially the short data sequence of vibration signal analysis;2.The cosine superposition method was proposed, and the improved HHTalgorithm is verified by Matlab simulation accuracy and rapidity;3.According to the statistics law, and puts forward the segmentation time window technique, based on the extreme value point density by segmented time window willhave a piece oscillation data according to extreme value point density and the EMDdecomposition again, to be able to aliasing modal decomposition out effectively.Finally, put forward the complete improved HHT algorithm, and through thealgorithm to analysis the tie line power oscillation data from angle fluctuation and fourplants system,the analysis results show that the improved HHT algorithm caneffectively restrain the endpoint effect and modal aliasing phenomenon, operation speedand real-time performance can meet the needs of the monitoring analysis of powersystem low frequency oscillation.