OPTIMUM INPUT SIGNALS FOR PARAMETER-ESTIMATION IN SYSTEMS DESCRIBED BY LINEAR INTEGRAL-EQUATIONS

被引:0
|
作者
RAFAJLOWICZ, E [1 ]
机构
[1] WROCLAW TECH UNIV,WYBRZEZE WYSPIANSKIEGO 27,PL-50370 WROCLAW,POLAND
关键词
D-Optimality integral equation description; Experiment design; Information matrix; Numerical algorithm;
D O I
10.1016/0167-9473(90)90067-R
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
The paper deals with the problem of optimum experimental design for estimation of parameters, which appear in the kernel of an integral equation describing a physical process. The autocorrelation function of a spatial random field acting as an input (stimulus) is the decision variable in the considered problem. Under integral constraints imposed on decision variables results parallel to the optimum experiment design theory of Kiefer and Wolfowitz are presented. In particular, necessary and sufficient optimality conditions are derived and a computational algorithm is proposed. © 1990.
引用
收藏
页码:11 / 19
页数:9
相关论文
共 50 条