摘要
应用了一种求解Lyapunov方程的新方法,称之为特征值方法。首先从大型矩阵的krylov子空间法降阶开始,再假设矩阵A是可以被对角化,就可以用A的特征值分解式来代替A和AT,由此变换后Lyqpunov方程就容易的求解了。最后利用简单的线性变换求得原来方程的解。
In this paper we use a new numerical method for solving Lyapunov equations,which is called the ei-genvalue decomposition method. At fist we assume that A can be diagonalized, so we can replaced A and A^T by their eigenvalue decompositions, then the transfomed Lyapunov equations can be solved easily. Finally we can get the solution of original Lyapunov equations by a simple linear transformation.
出处
《中州大学学报》
2008年第2期103-105,共3页
Journal of Zhongzhou University
基金
浙江省自然科学基金(Y107112)
关键词
LYAPUNOV方程
特征值分解
模型降阶
子空间
算法
Lyapunov equation
eigenvalue decomposition
model reduction
krylov-subspace
Arnoldi algorithm
