Scattered data interpolation simulink

Hi,

I tried to use a lookup table in a simulink project. However my data is not monotonically increasing. So it gave me errors for breakpoints. It is not very clear how to implement lookup a table for scattered data (1 dependant + 2 independant variables).

So I reverted to put a MATLAB function block to use scatteredInterpolant() in the function. It worked when I put it in a matlab script but not in a simulink MATLAB function. Here is the error for the simulink:

 %code
 Undefined function or variable 'scatteredInterpolant'.
 Function 'MATLAB Function' (#36.1523.1572), line 34, column 15:
 "scatteredInterpolant(P_ent_mod,D_ent_mod,E_s_mod)"
 Launch diagnostic report.

I want to specify that scatteredInterpolant worked well in a script but not in the simulink function block

My scattered model data are 3 .txt files which I import in the workspace in 3 column variables (no time dependency).

My questions are:

- How to do a lookup table for scattered data from 3 variables (no time dependancy and not monotonically increasing)?

- How to use the scatteredInterpolant() in a simulink MATLAB function block?

- Is ther other alternative to do interpolation in simulink than lookup tables with scattered data?

Thank you very much,

Joel Lapointe

[Christopher Dadswell]

Hi Joel,

Did you find a solution to this? Currently having a similar issue.

many thanks, Chris

Answer 1

[yang xiang]

write a script like this and execute:

a=xlsread('data.xlsx');
x=a(:,1);
y=a(:,2);
z=a(:,3);
F=scatteredInterpolant(x,y,z);
save('inp.mat','F');

then mask the block in simulink and add

load('inp.mat')

in the initialization tab.

at last, use interpreted matlab function block in simulink. use matlab funtion like this:

F(u(1),u(2))

I hope this helps you.

[Dave Hall]

I see the "scatteredInterpolant" "inp" in my 'inp.mat' file, and I get no errors from the initialization command in my mask, but I get the error "Undefined function or variable 'inp'". Any ideas?

[Dave Hall]

Ah, nevermind. I wasn't using the interpreted matlab function block. Now it works.