Re : PID Autotuning design Ziegler-Nichols

[cdm38]

Bonjour,

J'utilise actuellement le VI "PID Autotuning Design.vi"  avec la méthode de Ziegler-Nichols en boucle ouverte pour obtenir les paramètres PID de mon système à réguler.

En observant de plus près le code implémenté dans le toolkit, je me demandais pourquoi le VI "PID Structure Conversion.vi" est utilisé en mode "Série" puisque les paramètres de sortie du VI "PID Table (Modified Ziegler Nichols).vi" sont déjà K,Ti en s et Td en s.

Selon moi, il faut utiliser le VI "PID Structure Conversion.vi" en mode "Academic (Standard)" pour rester cohérent avec les équations des structures PID décrites dans l'aide NI.

Le résultat est certes peu différent mais différent quand même selon les 2 modes "Série" ou "Academic".

Merci pour votre aide

[Paolo_P]

Hello,

You post your question in the wrong forum. This one is an English speaking forum.

Please post it in the French speaking forum or rewrite this one in English.

Thanks

[nathand]

I can't answer the question but I can at least offer a translation:

"Hello,

I'm using the VI "PID Autotuning Design.vi" with the Ziegler-Nichols method in open-loop mode to obtain PID parameters for the system I need to control.

Looking at the code in the toolkit more closely, I'm wondering why the VI "PID Structure Conversion.vi" is used in Series mode since the parameters coming out of the VI "PID Table (Modified Ziegler-Nichols).vi" are already K, Ti (in seconds), and Td (in seconds).

In my view, the "PID Structure Conversion.vi" should be used "Academic (Standard)" mode to stay consistent with the structure of the PID equations described in the NI help.

The results are of course only slightly different, but different nonetheless, depending on the mode Academic or Series.

Thanks for your help."

[cdm38]

Thanks for your translation, so :

Hello,

 I'm using the VI "PID Autotuning Design.vi" with the Ziegler-Nichols method in open-loop mode to obtain PID parameters for the system I need to control.

Looking at the code in the toolkit more closely, I'm wondering why the VI "PID Structure Conversion.vi" is used in Series mode since the parameters coming out of the VI "PID Table (Modified Ziegler-Nichols).vi" are already K, Ti (in seconds), and Td (in seconds).

In my view, the "PID Structure Conversion.vi" should be used "Academic (Standard)" mode to stay consistent with the structure of the PID equations described in the NI help.

The results are of course only slightly different, but different nonetheless, depending on the mode Academic or Series.

Thanks for your help

[Paolo_P]

I'm actually working on it.

I'll tell you when i will find something new.

Thanks

[AndyClegg]

This forum topic gives some definitive statement about the structures and units used by the z-n autotuning VIs.

Hope this helps,

[cdm38]

Hi,

I begin to understand thanks to the post.

In fact, all algorithms developped in the "PID and Fuzzy Logic Toolkit" use the "academic" PID structure.

If you want to use Ziegler-Nichols method for autotuning, you will see in litterature that the tables for calculating PID parameters are designed for "series" PID structure.

That's why NI use "PID conversion structure.vi" to convert PID in "series" into the form "academic" because the all algorithms PID VIs expect this structure.

Is it right? If it is, there's no problem in NI code.

Thanks for your help