Technical Forum

STEP 7 Additional Software

Control value generated by PID with OB35 cyclic achive 0

Thread Starter: martinpa   Started: 6/14/2012 6:23 PM   Replies: 8

« Previous thread Next thread »
Page 1 of 1 (9 items)
  6/14/2012, 6:23 PM
Joined 6/14/2012
Last visit: 6/21/2012
Posts: 5
Rating:
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
I have a problem with control value generated by PID in OB 35 block. Both OB35 and FB41 blocks have set interrupt (cycle) at value 5s494ms.  Could you tell me whether the blue control course is correct (see attachment)? I realized Smith predictor and controled heather.

Attachment: controlValueProblem.jpg  (81 Downloads)



=== Edited by MPach @ 6/14/2012 6:24 PM [GMT ] ===



Top
  6/14/2012, 11:02 PM
Joined 1/28/2009
Last visit: 10/31/2014
Posts: 4716
Rating:
Rated: Excellent [4.22 out of 5 / rated 702 time(s)]. (702) 
Rated: Excellent [5 out of 5 / rated 1 time(s)]. (1) 
martinpa wrote:
I have a problem with control value generated by PID in OB 35 block. Both OB35 and FB41 blocks have set interrupt (cycle) at value 5s494ms.  Could you tell me whether the blue control course is correct (see attachment)? I realized Smith predictor and controled heather.


Hello  martinpa,

Smith predictor,We have seen that before here with SIMATIC
Products:

Smith-predictor for processes with long delay time


To tell you what is exactly wrong with your "Set up"
, we need to see your codes.From my point of view,Short
(im)pulses on a  control valve is not logical to achieve
a control strategy and make the service life of the actuator
shorten.


I just concluded  those points from the attachment provided.

Hope this helpsRnf:lblSmileyTooltip_header

Best regards,
Hamid Hosseini
Top
Thank you for this post from:
  6/15/2012, 12:07 PM
Joined 6/14/2012
Last visit: 6/21/2012
Posts: 5
Rating:
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Hello Hamid,
thank you for your answer.
Following your advice, I wrote a program on a pdf.
I scaled input range of 0 to 6912 in the range 0 to 100% and the output from 0 to 100% of range 0 to 6912.
The reference value is in the range from 0 to 100%.
OB35 block and function block FB41 are called cyclically every 5s494ms.
I tested my code (function blocks - inertia model and dealy model) in matlab
and it was correct.

Best regards,
Marcin Pach.

Attachment: OB1 and OB35.pdf  (90 Downloads)



=== Edited by martinpa @ 6/15/2012 12:20 PM [GMT ] ===


=== Edited by martinpa @ 6/15/2012 12:08 PM [GMT ] ===



Top
  6/15/2012, 8:21 PM
Joined 1/28/2009
Last visit: 10/31/2014
Posts: 4716
Rating:
Rated: Excellent [4.22 out of 5 / rated 702 time(s)]. (702) 
Users with special expertise supervising conferences
Rated: Excellent [5 out of 5 / rated 1 time(s)]. (1) 
Hello  again,
Please check if it possible to clarify following points:

1- In which mode do you use FB41.Normally,P_SEL
is activated.What about I-SEL and D_SEL.If you do
not want to use them simply put "FALSE" operand
instead of input contacts.If you use inputs to select
or de-select I_Part and D_Part in operation,please
specify!

2-Why your "cycle" of FB41 does not match to the
real cycle of OB35(Normally T#100ms)?Why your
selected cycle is too long?

3-Provide us the archive of the project if possible and
a brief about handling algorithm based on your design.





Best regards,
Hamid Hosseini



Top
Thank you for this post from:
  6/17/2012, 3:06 PM
Joined 6/14/2012
Last visit: 6/21/2012
Posts: 5
Rating:
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Hello  again,
thank you for your answer.
Cycle time due to the time constant (21.9778) divided by the value 4.
Could you tell me how do I properly set the PID of my settings: Gain 1.5; Ti 40s; Td 11s
On Monday will send you my project.

Best regards,
Marcin Pach



=== Edited by martinpa @ 6/17/2012 3:26 PM [GMT ] ===


Top
  6/17/2012, 4:23 PM
Joined 1/28/2009
Last visit: 10/31/2014
Posts: 4716
Rating:
Rated: Excellent [4.22 out of 5 / rated 702 time(s)]. (702) 
Users with special expertise supervising conferences
Rated: Excellent [5 out of 5 / rated 1 time(s)]. (1) 
martinpa wrote:
Hello  again,
thank you for your answer.
Cycle time due to the time constant (21.9778) divided by the value 4.
Could you tell me how do I properly set the PID of my settings: Gain 1.5; Ti 40s; Td 11s
On Monday will send you my project.

Best regards,
Marcin Pach

Hello  again,
Cycle in FB41 means what is the time between 2 successive
calls of it.The easiest way to avoid any error for that is using
this codes :


If you need bigger cycle time(default 100ms for OB35) ,
so go to HW config and define that for OB35.

and this is for your 2nd question:






Hope this helpsRnf:lblSmileyTooltip_header

Best regards,
Hamid Hosseini



=== Edited by hdhosseini @ 6/17/2012 8:28 PM [GMT ] ===


Top
  6/17/2012, 7:31 PM
Joined 6/14/2012
Last visit: 6/21/2012
Posts: 5
Rating:
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Thanks Hamind. I'll check your advice tomorrow in the lab.

Best regards,
Marcin Pach
Top
  6/19/2012, 12:58 PM
Joined 6/14/2012
Last visit: 6/21/2012
Posts: 5
Rating:
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Hello  again,
I carried out two tests.
1. Option D was turned off (only a PI controller) and the control after reaching the set point remains constant.
2. Constant decrease in Ti (PID works) meant that control after reaching the setpoint remains constant.

So option D introduces nervousness into the control system.

Best regards,
Marcin Pach
Top
  6/22/2012, 12:12 PM
Joined 1/28/2009
Last visit: 10/31/2014
Posts: 4716
Rating:
Rated: Excellent [4.22 out of 5 / rated 702 time(s)]. (702) 
Users with special expertise supervising conferences
Rated: no ratings [0 out of 5 / rated 0 time(s)]. (0) 
Hello  martinpa ,

For an overview of 3-term PID controllers,please check
THIS link.Also ,this topic for step response of a casual
differentiator, D part accelerate controller response while
a jump in error (e.g due to step change in set point) and fade
away.

diff function of modular PID

martinpa wrote:
2. Constant decrease in Ti (PID works) meant that control after reaching thesetpoint remains constant.

This would not be an "always True" statement,it depends
on process dynamic and disturbanceRnf:lblSmileyTooltip_  think!Effect of Ti on
Control value discussed here:

wrote:

Please check the attachment for step response of a PI
controller which is exactly used by SIEMENS in FB41.
This is extracted from here and manual introduced in last
post.
FB41 needs gain(KC in curve) and reset timeTi (T1 in curve).
Please note that term 'repeats/minutes' which is widely used
in DCS systems stands for result of (gain/Ti) or (KC/T1).With this
factor (as shown in attachment), the integral  mode tries to  repeat
proportional correction (h(t)) by the end of first integral time.The h(t)
is added by gain if error is consatnt and each integration time passes.

Note, the shorter the integration time more propertional correction  is
repeated (more repeats / minutes).


Attachment: PI_REG.zip 



Hope this helpsRnf:lblSmileyTooltip_header
Best regards
Hamid Hosseini
Top
Page 1 of 1 (9 items)
Actions