基于双耦合Duffing振子列与FSWT的次同步振荡模态辨识

发布时间：2018-06-15 02:40

本文选题：次同步振荡 + 模态辨识　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：考虑到我国的能源与负荷分布不均问题,大规模的远距离输电已经成为我国解决此问题的一种行之有效的手段。但是,远距离输电的大规模应用同样使得次同步振荡问题成为威胁电力系统稳定的重要隐患。近年来,我国已多次发生了次同步振荡现象。在对次同步振荡的分析中,基于量测信号的模态辨识一直是一个重要的研究方向。本文旨在结合Duffing理论与频率切片小波变换(FSWT)实现次同步振荡模态辨识,该算法可以准确地辨识出发生次同步振荡的模态,并对噪声有较好的免疫效果。本文首先考虑到经典Duffing振子理论在辨识信号方面的缺陷,提出了利用双耦合Duffing振子列辨识次同步振荡的频率。对复合衰减信号和次同步振荡仿真信号在含噪和不含噪情况下进行辨识,辨识结果表明:双耦合Duffing振子列是一种具有强噪声免疫性、可以准确辨识次同步振荡信号频率的辨识算法。然后,本文根据辨识出的次同步振荡频率,利用FSWT实现了对该频率分量信号的拟合。对拟合出的信号分量进行希尔伯特变换即可得到信号分量的衰减系数和阻尼比等参数。利用该算法对复合衰减信号和振荡信号进行拟合,结果表明FSWT是一种可以准确拟合信号分量的方法,且相较Prony辨识方法有更好的噪声免疫性。最后本文提出了利用双耦合Duffing振子列与FSWT的次同步振荡模态辨识算法。该算法结合了双耦合Duffing振子列对频率辨识的准确性与FSWT可以辨识出模态衰减系数和阻尼比的优势,实现了对次同步振荡信号的模态辨识。通过该算法对发生了次同步振荡的多机系统振荡信号与电厂实测次同步振荡信号的辨识,验证了该算法是一种对噪声有较好免疫性、可以准确辨识出次同步振荡振荡模态的算法,经进一步完善具有工程应用价值。
[Abstract]:Considering the uneven distribution of energy and load in our country, large-scale long-distance transmission has become an effective means to solve this problem in our country. However, the large-scale application of long distance transmission also makes the sub synchronous oscillation an important threat to the stability of the power system. In recent years, China has taken place many times. Synchronous oscillation phenomenon. In the analysis of sub synchronous oscillation, modal identification based on measurement signal is always an important research direction. This paper aims to realize the subsynchronous oscillation mode identification with Duffing theory and frequency slice wavelet transform (FSWT). This algorithm can identify the modes of the initial synchronous oscillation and the noise. In this paper, we first consider the defects of the classical Duffing oscillator theory in the identification signal, and propose a double coupling Duffing oscillator to identify the frequency of subsynchronous oscillation. The identification of the compound attenuation signal and the subsynchronous oscillation simulation signal in the noise and non noise conditions is carried out. The identification results show that the dual coupling Duff is two coupling. The ing oscillator column is an identification algorithm with strong noise immunity and can accurately identify the frequency of subsynchronous oscillation signal. Then, based on the identified subsynchronous oscillation frequency, the fitting of the frequency component signal is realized by using FSWT. The signal component can be attenuated by the Hilbert transform of the fitted signal component. The coefficients and damping ratios are used to fit the compound attenuation signal and the oscillating signal. The results show that FSWT is a method to accurately fit the signal components and has better noise immunity compared with the Prony identification method. Finally, the subsynchronous oscillation mode of the dual coupled Duffing oscillator column and the FSWT is proposed. The algorithm combines the accuracy of frequency identification with the dual coupled Duffing oscillator and the advantage of FSWT to identify the modal attenuation coefficient and damping ratio, and realizes the modal identification of the sub synchronous oscillation signal. It is proved that the algorithm is a kind of algorithm that has good immunity to noise and can identify the oscillation mode of sub synchronous oscillation accurately, and it has the value of engineering application after further improvement.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM712

【参考文献】