PID

[franz2]

Hallo,

ein Kollege schreibt gerade an einem LabVIEW- Programm und benutzt dort einen PID vi. Er hat allerdings nicht so viel Ahnung von der Regelungstechnik, daher nun auch seine Frage. Bei dem VI kann ein Anschluss linearity mit dem PID vi verbunden werden. Was hat es mit diesem Eingabefeld auf sich und was bewirkt es?

Leider kann ich die Bedeutung dieses Felds auch nicht genau erklären, daher poste ich diesen Eintrag.

[DOC (-) doc]

Could you please ask agin in english

[shoneill]

Approximate translation:

A colleague of his is using a PID vi and doesn't know too much about PID. There's an input variable called "linearity" and he's asking if someone could explain how this fits into the grand scheme of things, so to speak.

Hope this helps

Shane.

[Dr._Imad]

Thanks for the translation! Here is my answer in English

The linearity term appears as an input in the PID-advanced vi. It has a range from 0 to 1.

It's effect is to make the effective proportonal gain dependent on the error, which is the difference between the set point and the physical readout.

If set to 1 (default), then the proportional gain is Kc (no change)
If set otherwise, then the effective prop. gain will be dependent on the error, actually it inceases with the error.
Example, if L=0.2, then prop.gain=0.2Kc+ a term dependent on error and L
and the miniumm prop gain is going to be 0.2 Kc


If there are no sudden changes in the set point, set L=1 (or simply use the PID.vi (not the advanced)
If there are sudden changes in the set point, set L appropriately

[franz2]

Thanks for your answer.But I don´t understand the following sentence you wrote:
Example, if L=0.2, then prop.gain=0.2Kc+ a term dependent on error and L
and the miniumm prop gain is going to be 0.2 Kc. Can you explain it more detailed, please?

Thanks


P.S.: Is someone here, who can explain it in German?

[Dr._Imad]

The control signal = e*Kc + Integral(e) dt * KI + Kd de/dt
where Kc is the proportional gain
KI is the Inegral gain, and Kd is the Integral gain.


If L=1, the above equation is used .

If L is not equal to 1, then the proportional error contribution is changed. instead of being e*Kc, it is now changed to e*Kcmodfied. In effect, The gain multiplying the error e is modified from Kc to
Kcmodfied = Kc* [L + (1 - L) * abs(e) / sprng]

where sprng is the set point range. If the set point is fixed, or if it changes slowly (), then set L=1, and the modification is ignored. Kcmodfied=Kc

If L=0.2, the
Kmodfied=0.2Kc + the other (positve term that depend on L and the ratio of error to SetPointRange).
If the error =0, then Kcmodfied=0.2Kc, and this is the minumum value of the modfied proportional gain

[shoneill]

Franz,

I would translate if I understood it! Sorry, I'm not a PID expert at all and have the feeling I'd mix up too much if I started translating something I don't really understand.

Sorry.

shane.

[franz2]

Thanks, for that detailed explanation of the problem. It would be nice, if someone could explain it in German, because me and my friend don´t really understand it in English.