smith predictor with current temperature as input for Transfer Function

[@JuMos]

Hi,

I have implemented a smith predictor to control a temperature, but it does not work well. My System shows pretty much a PT1 behaviour with delay. But in the Smith Predictor, even if their is no heating, because the desired temperature is lower than my real temperature, the output of the simualted PT1 is never negative. So the model of my System (Transfer Function with PID) does not forcast the decline of the temperature. Is it possible to give the current temperatur as an input (like an initial condition) to the Transfer Function?

[Julika]

Hi,

this is Julika from National Instruments.

thank you for your detailed problem discription.

I did some research for you. Did you try tu use the "PID Autotuning (Temperature) VI"?

See the following link:

PID Autotuning (Temperature) VI - LabVIEW 2014 Help - National Instruments
http://zone.ni.com/reference/en-XX/help/371361L-01/lvpid/pid_autotuning_temperature/

Please attend especially the "measured temperature" input.

Please also check the example VI in LabVIEW: Autotuning Smith Predictor.vi. The above mentioned VI is also used in this example.

Best regards,

Julika Spitz

Applications Engineer
National Instruments