1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Adaptive PID controller implementation

  1. Jun 10, 2015 #1

    I am trying to implement a simple adaptive PID controller paper. However, I am having some problems which I could not find out due to what..

    Paper is the following;

    Quite simple. Using the Tagaki-Sugeno fuzzy system, in accordance with the error percentage, P and I parameters are adjusted. I implemented everything conforming to the paper. However, my control signal blows up immediately.

    I have been looking for days where I am doing wrong but could not find any mistake.

    I would really appreciate if you check the simulation files and make comments.

    Attached the simulation files.
    P.S. initialize.m must be run before running the simulink model.

    Thank you.

    Attached Files:

  2. jcsd
  3. Jun 11, 2015 #2


    User Avatar
    Gold Member

    what do you mean by your control system blows up. please provide screen shots or a more detailed explanation
  4. Jun 11, 2015 #3

    What I mean is, given reference signal is not tracked as expected.

    Here is the reference signal. Just a square wave.

    And the output of the plant as the response (I just limited it to 1 second, it goes with the same fashion up to 10 seconds)

    So the tracking is way out of the reference signal.
  5. Jun 12, 2015 #4


    User Avatar
    Gold Member

    post a screenshot of your code and the model
  6. Jun 12, 2015 #5
    All the files related to simulation were already attached my first post but anyways here's the screenshot of the model


    Here's the function that runs inside the interpreted matlab function block

    And those are the initializations;

  7. Jun 13, 2015 #6


    User Avatar
    Gold Member

    Why do you set the derivative gain to 1? try setting it to zero
    also you initial values for your P and I values should not be zero.

    try using zeiger-nicolas to set the initial values, and strip the adaptive part.
    once you get that working, add the adaptive portion
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook