通过轨迹特征根分析时变振荡特性

Trajectory Eigenvalues Analysis Time Variant Oscillation Characters

电力系统受扰轨迹具有非平稳振荡特性,无法由平衡点特征根反映。为此,提出受扰轨迹振荡模式的时变性概念,给出瞬时阻尼的计算式;采用小波脊方法获取轨迹各窗口内的特征根,构成振荡模式的时间序列,以便研究扰动大小和系统时变性对低频振荡的影响。小波变换利用滑动的伸缩窗口在多个尺度空间分析信号的时频特性,有较好的频率适应性。通过追踪小波变换的极大值点可以克服Prony和傅里叶算法不适用于时变系统的困难,反映时变振荡特性。指出由于小波变换反映的是窗口内的平均振荡特性,窗口过宽会掩盖其时变性,因此不能用于时变性太强的场合。

The disturbed trajectories of power system usually are of non-stationary oscillation characteristics, which can not be represented by equilibrium point eigenvalues. Therefore, the concepts of time-variant oscillation modes and trajectory eigenvalues are put forward. The way to calculate the instantaneous frequency and damping factor is also given. Time sequence of oscillation modes, composed of the eigenvalues in sliding windows along the trajectory and derived from wavelet ridge methods, clearly shows the influence of the non-linearity and time-variation on low frequency oscillations. By tracing the local maximum points of wavelet transform, this method overcomes the inapplicability of the Prony and Fourier algorithms to time- variant systems. However, the wavelet transform can not deal with fast time-variation eases because it treats the dynamics within a window as a stationary oscillation. For such cases, time sequence of trajectory eigenvalues at a series of moments should be developed.