Can vpasolver work in simulink model?

9 Ansichten (letzte 30 Tage)
Paris Pasqualin
Paris Pasqualin am 3 Dez. 2020
Kommentiert: Paris Pasqualin am 14 Jun. 2022
Hello to everyone,
I wanted to create a function block on simulink with outputs calculated from the solution of vpasolver. When I run the simulink model I get an error saying
"Function 'syms' not supported for code generation. Function 'MATLAB Function' (#35.101.118), line 3, column 1: "syms Qp cp Qb cb" Launch diagnostic report."
Does this mean that vpasolver can't work on simulink? If yes, could there be a way to use it? Beneath it's the function I'm referring to.
Best regards,
Paris
function [Qp,cp,Qb,cb] = fcn(Q_in,c_in,r,P,i,T)
R = 0.0832; % Universal gas constant [L*atm/(K*mol)]
syms Qp cp Qb cb
S = vpasolve(r == 1-cp/c_in, Q_in == Qp+Qb, Q_in*c_in == Qp*cp + Qb*cb, P/(i*R*(T+273.15)) == cb-cp, [Qp cp Qb cb]);
Qp = S.Qp;
cp = S.cp;
Qb = S.Qb;
cb = S.cb;

Akzeptierte Antwort

Uday Pradhan
Uday Pradhan am 16 Dez. 2020
Bearbeitet: Uday Pradhan am 16 Dez. 2020
Hi,
I think a possible workaround to this would be to do the symbolic computations in a separte function script and call that function from your block as an extrinsic function. Example: Define 'fcn.m' that contains all the computations:
function [qp,cp,qb,cb] = fcn(Q_in,c_in,r,P,i,T)
R = 0.0832; % Universal gas constant [L*atm/(K*mol)]
syms Qp cp Qb cb
S = vpasolve(r == 1-cp/c_in, Q_in == Qp+Qb, Q_in*c_in == Qp*cp + Qb*cb, P/(i*R*(T+273.15)) == cb-cp, [Qp cp Qb cb]);
qp = double(S.Qp);
cp = double(S.cp);
qb = double(S.Qb);
cb = double(S.cb);
end
Then in the MATLAB function block call:
function [Qp,cp,Qb,cb] = fcn1(Q_in,c_in,r,P,i,T)
coder.extrinsic('fcn');
Qp = 0;
cp = 0;
Qb = 0;
cb = 0;
[Qp,cp,Qb,cb] = fcn(Q_in,c_in,r,P,i,T);
I hope this helps!
  4 Kommentare
Gordon
Gordon am 13 Jun. 2022
I get the error that numvars = 0. So it's as if still not evaluating variables.
Paris Pasqualin
Paris Pasqualin am 14 Jun. 2022
Hi Gordon,
You actually do not need the vpasolve for simulink. You run the vpasolve once without spacifing any of the parameters and you should get equations that correlate the output parameters with the input ones. In my case I have:
Cp1 = (Cf*(156250*A1*Ps + 156250*A1*Lp*P - 156250*A1*Lp*P*sigma - 13*Cf*Lp*T*i1*sigma + 13*Cf*Lp*T*i1*sigma^2))/(13*Cf*Lp*T*i1*sigma^2 - 13*Cf*Lp*T*i1*sigma + 156250*A1*Ps) - (24414062500*A1^2*Ps^2 - 317382812500*A1*Ps*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) + 24414062500*A1^2*Lp*P*Ps - 169*Cf^2*Lp^2*T^2*i1^2*sigma^3 + 169*Cf^2*Lp^2*T^2*i1^2*sigma^4 + 26406250*Cf*Lp*T*i1*sigma*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) - 26406250*Cf*Lp*T*i1*sigma^2*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) + 2031250*A1*Cf*Lp^2*P*T*i1*sigma + 4062500*A1*Cf*Lp*Ps*T*i1*sigma^2 - 2031250*A1*Cf*Lp^2*P*T*i1*sigma^2 - 2031250*A1*Cf*Lp*Ps*T*i1*sigma)/(26*Lp*T*i1*sigma*(13*Cf*Lp*T*i1*sigma^2 - 13*Cf*Lp*T*i1*sigma + 156250*A1*Ps));
Cp2 = (Cf*(156250*A1*Ps + 156250*A1*Lp*P - 156250*A1*Lp*P*sigma - 13*Cf*Lp*T*i1*sigma + 13*Cf*Lp*T*i1*sigma^2))/(13*Cf*Lp*T*i1*sigma^2 - 13*Cf*Lp*T*i1*sigma + 156250*A1*Ps) - (24414062500*A1^2*Ps^2 + 317382812500*A1*Ps*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) + 24414062500*A1^2*Lp*P*Ps - 169*Cf^2*Lp^2*T^2*i1^2*sigma^3 + 169*Cf^2*Lp^2*T^2*i1^2*sigma^4 - 26406250*Cf*Lp*T*i1*sigma*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) + 26406250*Cf*Lp*T*i1*sigma^2*((A1^2*Lp^2*P^2)/169 + (2*A1^2*Lp*P*Ps)/169 + (A1^2*Ps^2)/169 - (A1*Cf*Lp^2*P*T*i1*sigma^2)/1015625 + (A1*Cf*Lp*Ps*T*i1*sigma^2)/1015625 + (Cf^2*Lp^2*T^2*i1^2*sigma^4)/24414062500)^(1/2) + 2031250*A1*Cf*Lp^2*P*T*i1*sigma + 4062500*A1*Cf*Lp*Ps*T*i1*sigma^2 - 2031250*A1*Cf*Lp^2*P*T*i1*sigma^2 - 2031250*A1*Cf*Lp*Ps*T*i1*sigma)/(26*Lp*T*i1*sigma*(13*Cf*Lp*T*i1*sigma^2 - 13*Cf*Lp*T*i1*sigma + 156250*A1*Ps));
Then you use these equations in simulink.
Kind regards,
Paris

Melden Sie sich an, um zu kommentieren.

Weitere Antworten (0)

Kategorien

Mehr zu Loops and Conditional Statements finden Sie in Help Center und File Exchange

Tags

Produkte


Version

R2018b

Community Treasure Hunt

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

Start Hunting!

Translated by