+ Post New Thread
Results 1 to 5 of 5
  1. #1
    Member level 2
    Points: 325, Level: 3

    Join Date
    Nov 2017
    Posts
    46
    Helped
    0 / 0
    Points
    325
    Level
    3

    what are the value I should put in PID controller of simulink and PLECS?

    Hello everyone,

    As I am trying to control a DC/DC converter, I am bit confused about the parameters of the PID. so I will ask these stupid questions.
    1-in simulink, I find the the variable to fill in the PID(z) Controller as P. I , and D. here I am bit confused ,should I write Ku(ultimae gain) in the box P, Ti in box I and Td in box D or Kp, Ki and Kd.
    2-in PLECS, trhe same problem arised as in the PI COntroller I found in one of their demos,Ki and Kp, what are these value representing? is it Kp and Ki(Ki=kp/Ti) or just Ti?

    thnak you so much for clarifying me#
    cheers

    •   Alt7th February 2018, 16:21

      advertising

        
       

  2. #2
    Advanced Member level 4
    Points: 6,571, Level: 19

    Join Date
    Dec 2015
    Location
    Madrid, Spain
    Posts
    1,180
    Helped
    284 / 284
    Points
    6,571
    Level
    19

    Re: what are the value I should put in PID controller of simulink and PLECS?

    2-in PLECS, trhe same problem arised as in the PI COntroller I found in one of their demos,Ki and Kp, what are these value representing? is it Kp and Ki(Ki=kp/Ti) or just Ti?
    I am not aware of any PID controller block in PLECS, can you share how/where did you find it ?

    1-in simulink, I find the the variable to fill in the PID(z) Controller as P. I , and D. here I am bit confused ,should I write Ku(ultimae gain) in the box P, Ti in box I and Td in box D or Kp, Ki and Kd.
    When you open the PID(z) block you can see how it is implemented. You need to insert the latter option you mentioned.



    •   Alt7th February 2018, 17:56

      advertising

        
       

  3. #3
    Super Moderator
    Points: 15,912, Level: 30
    E-design's Avatar
    Join Date
    Jun 2002
    Posts
    2,014
    Helped
    503 / 503
    Points
    15,912
    Level
    30

    Re: what are the value I should put in PID controller of simulink and PLECS?

    I think the PID is available in the later versions of PLECS. It is not available in my old version.
    https://www.plexim.com/support/application-examples/187



    •   Alt7th February 2018, 18:19

      advertising

        
       

  4. #4
    Advanced Member level 4
    Points: 6,571, Level: 19

    Join Date
    Dec 2015
    Location
    Madrid, Spain
    Posts
    1,180
    Helped
    284 / 284
    Points
    6,571
    Level
    19

    Re: what are the value I should put in PID controller of simulink and PLECS?

    Quote Originally Posted by E-design View Post
    I think the PID is available in the later versions of PLECS. It is not available in my old version.
    https://www.plexim.com/support/application-examples/187
    The example you showed is called (at least in my version) "Buck Converter with loop gain Analysis". You can find it under demo models --> Power supplies.
    The "PID controller" block is simply a subsystem, not a block per se.



  5. #5
    Full Member level 6
    Points: 2,927, Level: 12

    Join Date
    Feb 2014
    Posts
    355
    Helped
    126 / 126
    Points
    2,927
    Level
    12

    Re: what are the value I should put in PID controller of simulink and PLECS?

    I'm not familiar with PLECS specifically but in my experience although the term "PID" is widely used there is nothing close to a standard implementation of "PID" or its coefficients (there are many ways to arrange the PID blocks, scale the coefficients, etc).

    So the right approach is to decide what you want the PID to do in terms of its frequency response. Then use the AC/frequency domain simulation (which PLECS appears to have) to verify that your specific PID implementation and coefficients are doing that. Don't just trust the documentation and a stable closed loop also doesn't guarantee its doing what you want.


    As another aside while P, I and D blocks can be combined to meet almost any control requirement (and discrete PID blocks are currently my preferred digital control solution versus, say, biquad IIRs) I'm aware of no systems where a single P, I and D (two zeros and one pole) is an ideal compensation solution. So its perplexing to me that PID is presented as such a common and general control scheme.



--[[ ]]--