4D integral with exponential value

16 次查看(过去 30 天)
Hi,
I am trying to calculate this integral using indefinite integral functions.
1 1 / 1 1
/ / | / /
| | | | | 1
| | exp| | | ----------------------------------------------------------------------- dx dy
/ / | / / / 1 \ / 2 / 1 \2 \2
0 0 | 0 0 | ------------------------------ + 160 | | (w - z) + | y - x + - | |
| | / 2 / 1 \2 \2 | \ \ 2 / /
| | | (w - z) + | y - x + - | | |
\ \ \ \ 2 / / /
\
|
|
| dz dw
|
|
|
|
/
It is 4D integral. When I try to use indefinite integral function int Matlab cant calculate it.
int(int(exp(int(int(1/((1/((w - z)^2 + (y - x + 1/2)^2)^2 + 160)*((w - z)^2 + (y - x + 1/2)^2)^2), x, 0, 1), y, 0, 1)), z, 0, 1), w, 0, 1)
When I try to use integral2 function to calculate numerically, the z and w parameters are obstacles.
I cannot use 4D user defined integral function because there exp{} block between integrals.
Do you know how to solve that integral?

回答(2 个)

Torsten
Torsten 2017-3-21
编辑:Torsten 2017-3-21
Brute-force method:
delta=0.01;
w=0:delta:1;
z=0:delta:1;
x=0:delta:1;
y=0:delta:1;
func=@(x,y,w,z)1/(1+160*(((w-z)^2+(y-x+0.5)^2)^2));
for i=1:numel(w)
wloc = w(i);
for j=1:numel(z)
zloc = z(j);
f = 0;
for k=1:numel(x)-1
xloc_l = x(k);
xloc_r = x(k+1);
for l=1:numel(y)-1
yloc_l = y(l);
yloc_r = y(l+1);
f = f + 0.25*(func(xloc_l,yloc_l,wloc,zloc)+func(xloc_l,yloc_r,wloc,zloc)+func(xloc_r,yloc_l,wloc,zloc)+func(xloc_r,yloc_r,wloc,zloc))*delta^2;
end
end
fxy(i,j) = exp(f);
end
end
integral = 0.0;
for i=1:numel(w)-1
for j=1:numel(z)-1
integral = integral + 0.25*(fxy(i,j)+fxy(i,j+1)+fxy(i+1,j)+fxy(i+1,j+1))*delta^2;
end
end
"integral" should be the value of the integral you are looking for.
You may want to choose a smaller value for "delta" to get a better approximation.
With delta=1/300 I get a value of 1,166618 for your integral.
Best wishes
Torsten.

Mümin ÖZPOLAT
Mümin ÖZPOLAT 2017-3-22
编辑:Mümin ÖZPOLAT 2017-3-22
Thanks !
Seems like it is an implementation of Riemann Sum.
You use delta^2 because the integration is 2 layered and multiplicate it by 0.25 because you get average of 4 different result.
  1 个评论
Torsten
Torsten 2017-3-23
It's the two-dimensional trapezoidal rule, applied twice: once to the double inner integral, then to the double outer intergral.
Best wishes
Torsten.

请先登录,再进行评论。

类别

Help CenterFile Exchange 中查找有关 Numerical Integration and Differentiation 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by