用时间序列来判断电力系统中的混沌现象

由于混沌现象对初值、参数等的敏感依赖性，现有的基于电力系统的模型（方程）来研究混沌的方法，会由于建模时方程的形式、参数等误差带来可能的失误。为了弥补这一缺点，文中提出了直接利用对实际系统RS采样得到的时间序列的Lyapunov指数λi分析的方法来判断混沌的存在性。利用矩阵的特征值移动技术可以找到RS的伴随系统RS*，使RS* 的λ*i=λi+k(k为合适的正实数)，于是可以利用现有的计算时间序列Lyapunov指数的方法来实现RS的混沌判断。

USING TIME SERIES TO DETECT CHAOS IN POWER SYSTEMS

Troubles could be brought about in power system chaos study when the real systems are modeled by equations,which have inevitable errors in the initial conditions. forms and parameters. because chaos is sensitively dependent on them.To remedy the defects, this paper presents a new method to detect chaos directly by the time series sampled from the realsystem. The windowed Lyapunov exponents A, of the time series can indicate whether the real system (RS) is in chaos. RS'is an accompany system of RS. that can be constructed via the eigenvalues moving technique of matrix. such that A,' ~A,+k(k is a suitable positive number). to detect the stability of RS. This is due to the existing methods cannot directly computethe non-positive Al of a stable time series.