Research On Precise Identification Method For Fast Attenuate Component Of The Power System Low-Frequency Oscillation

The stable power supply of power system is vital for the national economy and social civilization. Stability is the key to the safe operation of power system, and low frequency oscillation is an important factor that affects the safe and stable operation of power system. With the growing scale of interconnected power system, the problem of low frequency oscillation is more and more serious, and even leads to the collapse of the whole power system. Therefore, it is of great significance to study some kind of identification algorithm that can deal with low frequency oscillation signal better which has various oscillation characteristics.Prony algorithm is a common analysis algorithm and important settling means of the power system low frequency oscillation. As a method of time domain of modal parameter identification, it can calculate the frequency, attenuate factor, amplitude and phase of the system by analyzing the measured oscillation data. Based on the analysis of the basic principles of the traditional Prony algorithm, a choice of strategy of the main parameters is given which affected the accuracy of the algorithm. Considering that the algorithm is very difficult to analyze the fast attenuate component of the low frequency oscillation mode and the fast attenuate component can be missed easily, an improved method of extracting rapid attenuate signal is proposed which is called Prony segment analysis method. This method divides the data acquisition time into two periods, analyze the second period firstly is used to calculate the slow attenuate component, and then analyze the first period for calculating the fast attenuate component. Numerical example proves that this algorithm can accurately extract fast attenuate component of low frequency oscillation mode, and also makes the slow attenuate component more accurate. This algorithm can obtain the low frequency oscillation signal.From the perspective of the analysis, the error of improved algorithm that is based on the segmentation of Givens transformation is discussed. Firstly, the sources and classification of the error are discussed. Then, the error of Givens transformation is analyzed in detail. Then combined with the specific examples, the error of the extracting modal information is compared between the Givens transformation subsection and the normal equation subsection method. The comparative results show that the identifying modal information error of the improved algorithm is smaller and the accuracy is better.