kringing model(random process value at given sample)

2 次查看（过去 30 天）

Chongqing Fan 2016-11-29

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/314467-kringing-model-random-process-value-at-given-sample

First, thanks for your attention for my question, my code is as follows:

 t = [0,1,2,2.5,5];
f=@(t)1.0417*t.^5-13.25*t.^4+59.792*t.^3-112.75*t.^2+75.167*t;
theta = 0.01; 
lob = 1e-5;
upb = 20; 
l = length(t);
for i = 1:l    
fun(i) = f(t(i)); 
end
dmodel = dacefit(t',fun',@regpoly0,@corrgauss,theta,lob,upb);
[g,dy,mse,dmse] = predictor(t',dmodel);

When i run the code, the mse and dmse is NAN, based on the DACE tool box, the mse and dsme should be two dimension. fit fun of kriging is y = f*b+e, where f is the basis function, and b is the unknown parameters, e is random process, finally, how to calculate the random process value at the given sample point? thanks again!