主要研究了一种隐式重新启动的Lanczos算法在模型降阶中的应用。分析了由这个算法得到的降价后的模型的一些性质,对于一个n阶稳定的线性时不变系统,模型降阶的思想是寻找一个m阶转换函数来近似原系统的n阶转换函数H(s),其中,n〉〉m,传统的krylov子空间方法仅仅产生一个不稳定的实现,并且在低频处的误差较大,本文所考虑的隐式重新启动的Lanczos方法,能较好的解决上述两个问题。
This Paper considers an implicitly Lanczos Algorithm that approximates a stable, linear transfer function f(s) of order n by one of order m, where n〉〉m, It is well known that obilqtie projections onto a Krylov subspace may generate unstable partial realizations. A second difficulty arises from the fact that Krylov subspace methods often generate partial realizations that contain nonessential modes. The method considered in this paper can remedy these situation.