Would a control system consisting of 2 PID controller one plant would be considered as an cascade controller??

And how come would a proper tuning method be?

As far i've googled it seems to me that only best method is to manually do it, one by one.

this is how my system looks like http://snag.gy/rJH2J.jpg


A cascaded controller simply has the output of one controller acting as the input to another.

A good example of this would be in a motor control application where you typically have a current loop controlling the current to the motor, over that you'd have a velocity loop where the output of that loop is the input to the current loop, and over that you'd have a position loop where the output is the input to the velocity loop.

Each loop can be tuned independently of the other by whatever method you fancy. Tune the inner-most controller first and then start going outwards.

| improve this answer | |
  • $\begingroup$ What kind of tuning method would you use, if ziegler nichols can't be used. $\endgroup$ – Control May 22 '14 at 21:25
  • $\begingroup$ @Control Why can't Z-N be used? $\endgroup$ – Guy Sirton May 22 '14 at 21:32
  • $\begingroup$ The system can't as far i can tell never become unstable $\endgroup$ – Control May 22 '14 at 21:44
  • 1
    $\begingroup$ @Control Then just set all the gains to infinity and you got yourself a great controller. Just kidding. If you're unable to get the system to become unstable by increasing the gains (e.g. of that inner loop) then there's something else going on that's unrelated to the cascade. I should add that it's easiest to see that this inner controller is "just" a regular non-cascaded controller. If you're unable to turn off the other controllers in the cascade you may need to set their gains low enough to be able to essentially see the inner controller in isolation. $\endgroup$ – Guy Sirton May 22 '14 at 21:46
  • $\begingroup$ Well.. my first approach was to isolate the inner loop, and then using z-n method, but since (according to root locus) it will not become marginally stable, i a bit lost on how i should aproach the situation. From what i've read is that a P controller would work for the inner loop and a PI for the outer, but since the output of the inner loop doesn't resemble the output of the system... i think something completely wrong. I've included a pic showing my control system, including the plant G(s), if it might help deducing the error. $\endgroup$ – Control May 22 '14 at 22:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.