How would one test the system response of the following example control loop

[labnoob]

Basically, I want to control the upstream pressure in a pipe, using a single setpoint PID loop on a valve. As the pressure changes up or down, upstream, I want to open or close a valve appropriately to maintain say 5kpsi.

I will used the advanced auotune PID loop, My instrument provides a 4-20mA, and my valve accepts a 4-20mA signal.

Would I be able to simulate this without taking the equipment to a live system? I have a simulated 4-20mA that I could inject into the DAQ and increase/decrease as required to simulate pressure changes, and the DAQ could give a 4-20mA signal to the valve.

But how would this all work in practice. Id like to see the valve respond, but unsure how I would be able to verify the valve response and upstream pressure

Any ideas or tips?

[mfitzsimons]

You should be able to simulate without equipment to get the PID response you are looking for.  The best place to state would be the LabVIEW example programs.  Use Help -> Find Example and search for PID.  I used this one to get the PID response I was looking for: C:\Program Files\National Instruments\LabVIEW 2011\examples\control\PID\Autotuning PID Online.vi

SOme systems respond quickly and others much slower.  Some systems can't deal with overshot and there are many design considerations.  You will need to tune the PID look for what is best for your system.If the system is straight forward and you can apply "rules" then consider the Fuzzy Logic controller.  I have had success with both but there is no magic answer for PID loops except trial and error.

[labnoob]

Hi mfitzsimons,

how would the PID take account of the physical valve response though?

is fuzzy logic in replace of PID, or used in line? Ive never experimneted with them before

thanks