Hello all,
As you can tell from my signature, I'm very new here, so please be patient.
I have a Simulink model of a HVAC fan controller with multiple plants/controllers in the system. The s-functions are templates found in Simulink's library. One controller in particular is intended to output a 7-element vector to a dmux. When I attempt to run the simulation, Simulink gives an error that "Output returned by S-function "...." in "...." during flag=3 call must be a real vector of length 7."
Doing a bit of digging - the vector length is fine, but it believes one variable is NaN. If I remove the variable from the output string, the simulation runs fine (not accurately, but it runs). If I allow the variable to output to the workspace (by removing a semi-colon) - the variable has a value. No doubt about it. If I shorten the simulation run time, and check the variable values at start-up, it still gives a positive, real number.
Some people have mentioned using a structure on it - which I attempted, but was stopped by Matlab, which stated I could not use a structure on a double.
I'm at my wits end - any ideas?
As you can tell from my signature, I'm very new here, so please be patient.
I have a Simulink model of a HVAC fan controller with multiple plants/controllers in the system. The s-functions are templates found in Simulink's library. One controller in particular is intended to output a 7-element vector to a dmux. When I attempt to run the simulation, Simulink gives an error that "Output returned by S-function "...." in "...." during flag=3 call must be a real vector of length 7."
Doing a bit of digging - the vector length is fine, but it believes one variable is NaN. If I remove the variable from the output string, the simulation runs fine (not accurately, but it runs). If I allow the variable to output to the workspace (by removing a semi-colon) - the variable has a value. No doubt about it. If I shorten the simulation run time, and check the variable values at start-up, it still gives a positive, real number.
Some people have mentioned using a structure on it - which I attempted, but was stopped by Matlab, which stated I could not use a structure on a double.
I'm at my wits end - any ideas?
