call different matlab function in for loop (SIMULINK)
1 次查看(过去 30 天)
显示 更早的评论
Dear community,
I'm having trouble getting a function to work, that calls a different beta function depending on the interval.
The getBeta_l function is created in a Matlab script, and has a suffix depending of the specific curvenumber.
How do I call the function and link it to my simulink variable s (curvelength)
manually it would look like this:
if (0 <= s)&& (s<= R_kpi/2)
beta = getBeta_1(s);
elseif (R_kpi/2 < s)&& (s<= R_k3pi/2)
beta = getBeta_2(s);
elseif
.
.
.
.
now I'm trying to do this in a for loop, but I can't figure out, how to implement it properly...
if (0 <= s)&& (s<= R_kpi/2)
beta = getBeta_1(s);
else
for q = 2:nuc %nuc = number of curves
if (R_k(2q-3)pi/2 < s)&& (s<= R_kpi(2*q-1)/2) %Interval
filename=['getBeta_l',num2str(q)]; %get function
beta=filename(s); %call function (s)
end
end
end
Thanks for your help,
Johannes
0 个评论
回答(2 个)
Udara De Silva
2022-1-27
Hi Johannes,
You can use Matlab 'run' command to do this as shown below:
if (0 <= s)&& (s<= R_kpi/2)
beta = getBeta_1(s);
else
for q = 2:nuc %nuc = number of curves
if (R_k(2q-3)pi/2 < s)&& (s<= R_kpi(2*q-1)/2) %Interval
filename=['getBeta_',num2str(q)]; %get function
run([filename, '(', num2str(s), ')']); % run function
beta=ans; % assign ans to output variable
end
end
end
Best Regards,
Udara
5 个评论
Udara De Silva
2022-1-29
编辑:Udara De Silva
2022-1-29
It seems that the error is thrown by Matlab Coder. So I assume although 'run' works in matlab it may not work in simulink. Can you try 'feval' instead of 'run'?
Walter Roberson
2022-1-29
discretize() to get an index number. Use it to index a cell array of function handles, and execute the resulting handle. No loops needed, and no if statements other than checking to be sure the input is within the bounds of discretization.
2 个评论
Walter Roberson
2022-1-31
Each time your q increases by 1, you move a "window" of width R_k further right and test to see if s is inside that window.
You could instead use
floor((s - SomeOffset)/Rk)
to directly calculate the appropriate q value. Discretization not needed.
You would have initialized
BH = {@getBeta_1, @getBeta_2, @getBeta_3 ... }
but first you would have put in an
assert(nuc <= SOME_LIMIT)
and then when you construct H, it will be either SOME_LIMIT or SOME_LIMIT+1 or SOME_LIMIT+2 entries (depending how you want to handle the boundary conditions on the two sides.) This assert() enforces that the index that gets used is no more than the maximum you configured for.
Once you have the index,
f = BH{q};
beta = f(appropriate_parameters);
Use real functions, not scripts for this purpose; Simulink cannot generate code for scripts.
另请参阅
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!