discretitization of the pid algorithm

Discussion in 'Homework Help' started by numericman, Feb 1, 2012.

  1. numericman

    Thread Starter New Member

    Dec 2, 2011
    21
    0
    I'm new to control.i'm trying to use numerical integration methods to obtain several discrete pid algorithms.I've just discretized the continuous time pid algorithm using euler forward integration,euler backward integration,trapezoidal and simpson methods.i now want to do a comparative performance analysis of the discrete pid algorithms obtained.please i need several case studies to analyze these discrete algorithm.please where can i get some direct and simple case studies? i've gone through several text but the things there are so complicated and confusing.thank you.
     
  2. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    The classical target for PI control (at least) is the first order lag plus pure delay [dead time] process model.

    The attached pdf may be of interest.
     
  3. numericman

    Thread Starter New Member

    Dec 2, 2011
    21
    0
    many thanks tnk.with your material provided further understanding of my project.sadly i've another problem which i've just posted.please have alook @ it.
     
  4. christin003

    New Member

    Feb 20, 2013
    3
    0
    I am also using simpson's discretization to change the continuous PI to the discrete PI however it results in three poles and one of them is unstable. Did you have any issues when you were discretizing? Also did you use simpson's 1/3 or 3/8 rule?
     
  5. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    The OP may not remember the answer to your question, given that the OP was posted over a year ago...

    You should start a new thread....

    What's wrong with the Bilinear Transform?
     
  6. christin003

    New Member

    Feb 20, 2013
    3
    0
    The transfer function for the system is discretized using bilinear transformation however I'm testing the robustness of the controller when its discretized using different methods. I started a new thread as you suggested :) thank you
     
Loading...