基于类噪声数据的电力系统低频振荡模式辨识方法及应用

低频振荡的监测对于电力系统的安全稳定运行是一个巨大的挑战。提出了基于类噪声数据的低频振荡模式在线辨识方法，该方法将类噪声PMU数据经过预处理后，以ARMA方法计算得到单测点低频振荡模式信息，然后通过聚类方法得到系统振荡模式信息。结合实际发生的一次低频振荡事故，通过比较扰动前和扰动过程中低频振荡模式差异判断振荡类型，并通过势能增量分布法予以验证。

Ambient Data Based Identification Method for the Low frequency Oscillation Modes of Power System and Its Utilization

The low frequency oscillation monitoring of power system is a huge challenge to its safe and stable operation. This paper presents an ambient data based identification method for the low frequency oscilation of power system. Preprocessing the PMU （ ambi-ent phasor measurement unit ） data from multiple PMU sites, the method can obtain the low frequency oscillation information at vari- ous PMU sites with ARMA （autoregressive moving average） calculation, and then derive the low frequency oscillation information modes of the power system with a clustering method. With an actual oscillation event in CSG （China Southern Power Grid）, the os-cillation type of the event is judged by comparing the estimated damping ratio before and during destabilization period, and verified by potential energy increment distribution method.