随机代数Riccati方程的条件数分析

Condition number analysis of the stochastic algebraic Riccati equation

对随机代数Riccati方程的条件数进行了研究,分别定义它的混合型、分量型及范数型条件数,并得到其显示表达式及上界.其次进行数值实验,验证了所提出的条件数能够给出比较好的误差界.

Conditional numbers can characterize the sensitivity of the solution to input data perturbation.In this paper,the condition numbers of stochastic algebraic Riccati equation are studied.The mixed,componentwise and normwise condition numbers are proposed,and their explicit expressions and easier computable upper bounds are derived.Finally,numerical experiments are carried out to show that the proposed condition numbers can give better error bounds.