Research Of Online Identification Of Low Frequency Oscillation In Power System Based On The Prony Algorithm

Online oscillation characteristic analysis algorithms are used to monitor online low frequency oscillations for control of wide area damping. This paper based on basic principle of Prony algorithm presents selecting strategies of main parameters. The sample data is preprocessed with improved wavelet soft-threshold technology for this algorithm is sensitive to noise, which increases signal-to-noise ratio and decreases influence of analysis results by the noise. As this technology has ineffective calculation,an improved Prony method for online analysis is proposed in this paper.The sampled signal is preprocessed first,then the identification of the practical order of the model and estimation of the parameters is improved significantly.The simulation results show that this algorithm is effective for identifying the order, improving the calculation speed and decreasing the influence of the noise.This method can satisfy online identification of LFO in power system.