PID Autotuning in LabVIEW

[MartinP]

I worked a lot with PID control from LabVIEW of various heaters and ovens like 15 years ago. My general impression of the Autotuning functionality back then was that, in addition to having one of the GUIs with the greatest potential for improvements I ever have come across in the LabVIEW world, the autotuning itself almost always failed. And yes, I used the special toolkit for PID control.

   My question now is if this have improved significantly in both respect the last decade, or if this still is a weak spot within the toolsets available from NI ?  Comments, anybody ?

Martin

[NPowl]

Hey Martin,

What was your concerns with the Autotuning function back then? I wasn't using LabVIEW back in those days, but I've had success with the PID autotuning. You have to establish a stable open-loop control system first (essentially manual tune it to roughly correct) before applying the autotuning, but it takes the rough estimation and fine-tunes it successfully. 

Cheers,

Nick

[MartinP]

Nick

The concern was that it simply didn't work. It seldom ended, giving me three new PID values to work with. And if it did, the three new values were no good. In addition, the user interface, see attachment, was nothing I would have served a paying customer.

I believe I had to implement the Cancel-functionality myself, if I remember correctly.

Martin

[NPowl]

Hey Martin,

I've been able to use the PID autotuning function (with the caveats in the earlier post) to success. I never used the Autotuning Wizard as in your screenshot, but did open it up and attached what the new UI looks like. Looks similar but a bit nicer with the NXG style by default.

Cheers,

Nick