• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 82 subscribers
  • Views 1357 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Decoupling of Id-Iq Components?

Felix Schinkel
Prodigy 210 points
Other Parts Discussed in Thread: CONTROLSUITE

Hello,

my next project will be controlling a pmsm with a F28x µC. In order to get started, i began to study some sample projects that are delivered with controlSUITE.

Here is my question:

In none of these projects is a decoupling of id and iq component. The output of the pi-controller goes directly into the ipark block an then into the space-vector gen. macro.

Is there a decoupling and i did not recognized yet, or is there really none? If second is true: why is there none? For having good dynamic behaviour, there should be a decoupling. 

  • 0
    Genius 5885 points

    Hi Felix,

    my guess is that while decoupling is an interesting concept academically, in practice the gain is quite minimal compared to the complexity it adds.

    Also in academia there is a lot of effort put in faster dynamic response, while in practice this is something that is sometimes actually avoided. While fast torque response sounds like something good, and it usually is, you really don't see reference values having a step form. Usually you put slope limiters to the reference inputs. This does make your system slower, but it reduces mechanical stress and audible noise.

    Regards, Mitja

  • 0 in reply to Mitja Nemec
    Prodigy 210 points

    Hi Mitja,

    thank you very much for your fast answer. I did not know that the pratical gain is that low. I have read much literature about that. All of them said, that it is very important to have such a decoupling network. But i think none of them have build such a controller. As always, there is a big difference between practice and theory.

    Felix

  • 0 in reply to Felix Schinkel
    Genius 5885 points

    HI Felix,

    Also another point about decoupling in order to clear possible confusion (I assume you already know this but it might be useful to somebody else).

    Some authors call the process of transforming current into two components (magnetizing and torque) decoupling. While this is technically correct it is not to be mistaken with decoupling d and q axis controller.

    The way usually (in practice and by TI) FOC is done, is that you transform current into magnetizing and torque component and send them through ordinary PI (or PID) controllers. This system works quite alright but you can see a disturbance in quadrature axis if you put a step change in direct axis and vice versa. The disturbance is there only at transients as current controller will track steady state reference value. It is this disturbance that you try to avoid when fully decoupling d and q axis controller. If you succeed q-axis should be completely oblivious to any transients in d-axis and vice versa. This is doable but as stated, not worth the effort.

    Regards, Mitja