PID Nightmare

You can have the PLC change the PID settings when the temperature gets 15 degrees from the setpoint.

I did this way back when.... http://www.plctalk.net/qanda/showthread.php?t=27555

I ended up using two PID's as suggested by John Gaunt in this thread, worked good but you need a lot of feedback as Peter is asking about, if you are trying to control blind and dont have enough feedback as to what is happening in the real world you will get mediocre control at best and if this is the case just use a open/close and forget the PID and I think this is the case as the OP can run the system manually just fine then that means either the PLC does not have the right information being fed back to it OR the correct feedback information is not available but either way it needs more work from a design aspect.
 
There are too many threads showing how not to approach tuning a system.
My first thought go to this William Thomson AKA Lord Kelvin quote.
“When you can measure what you are speaking about, and express it in numbers, you know something about it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts advanced to the stage of science.”

This is one my favorite quotes.

This means getting a model of the system to be controlled.



The problem is that system designers often are clueless and don't think about how the system will be controlled. The problem is that in 35+ years I have NEVER seen a system come with a documented transfer function. In other words, most mechanical engineers kludge their systems together and have no idea how it is going to work in reality.


Then they stick some poor unsuspecting/unkowing PLC guy with controlling it. Yeah, we know how that works out.
 
And when you point out the mechanical shortcomings, they ask why you don't fix that through some clever programming.
 
We have a few large jacketed kettles with heating and cooling controls. The big problem I see is the dead time associated with these type of controls because you are trying to heat/cool a large volume (I’m guessing op has a large vessel) and it takes time to do so. This is why you see the valve open to 100% for quite a while before closing as the PV reaches the SP. Feedback on the valve’s position probably won’t be of any use unless its response time is very high which I doubt. 14C overshoot is quite a bit. I’d lower the steam pressure some if heating time is of no concern. Second, as other suggested, fixed the output to 5% when you are at 15C below SP and let the PID control this or just use on/off 5% control for the last 15 degrees.
 
+1 for Split Range.

General Steps:
Process Identification Test
Process Model (Usually FODPT or SOPDT)
Ballpark PID parameters based on Process Model
Fine tune based on Trend (SP,PV and CV)

Autotune is an option with PIDE
 
We have a few large jacketed kettles with heating and cooling controls. The big problem I see is the dead time associated with these type of controls because you are trying to heat/cool a large volume (I’m guessing op has a large vessel) and it takes time to do so. This is why you see the valve open to 100% for quite a while before closing as the PV reaches the SP. Feedback on the valve’s position probably won’t be of any use unless its response time is very high which I doubt. 14C overshoot is quite a bit. I’d lower the steam pressure some if heating time is of no concern. Second, as other suggested, fixed the output to 5% when you are at 15C below SP and let the PID control this or just use on/off 5% control for the last 15 degrees.
I did a webinar using goto meeting years ago. I showed how to implement a Smith Predictor to compensate for the dead time.
On my YouTube channel my Smith Predictor and Sliding Mode Control is my second most view video
https://www.youtube.com/watch?v=uhLMyOlwCoM
 
That tool looks familiar; what was it called? I remember helping one of my boss' summer students with that many years back; it was pretty slick for its day; I had thought IDL was the bees knees until then.


[Update: d'oh! I had the window scrolled up; it's MathCad].
I have over 2500 Mathcad worksheets I have developed over 25+ years.


I think I could train 10 people to tune or re-tune all the systems in the US and Canada quickly. It is no longer necessary to be on site. I just need the data.
But don't worry. I will leave that for someone else to do.
 
Found two old threads, about tuning large heating systems, that may be relevant to this thread.

https://www.plctalk.net/qanda/showthread.php?t=104503


http://www.plctalk.net/qanda/showthread.php?t=4619


Cool.
The second thread is what encourage me to make the hotrod.zip in the misc section of the downloads section.


I think Ron Beaufort and I made a good complementary team. He would teach how to setup the PID on the PLC. I was never interested in that because Rockwell made it too complicated.
I would supply the control theory. We never stepped on each other's toes. We were in tune, in sync or complementary.



There is nothing new under the sun.
This is sad. The second thread is 17 years old. It seems that no one learns. I know there is always a new batch of rookies but they always seem to think they are the first to come across their problem.


I am 67 now and I sometimes wonder if people will ever learn.The same topics come up over and over again. Chances are what ever you are trying to do it has been done before. I always search first.
 
I am 67 now and I sometimes wonder if people will ever learn.The same topics come up over and over again. Chances are what ever you are trying to do it has been done before. I always search first.




It always cracks me up when a "new" query comes up here, and I poke at The Google, and the top useful responses are a few other threads on this very forum.


I mean, do some of us have a rare magical ability to select the right keywords, or are the OPs just not making the effort? I can't believe it's the former, and I don't like the way the latter thinks unkindly of the community, so maybe it's summat else.


But why does it feel like we're in that jokes-by-number story?
 
dbb, so jaded in such a short time, :ROFLMAO:
I'm am certainly no expert, but l play with a PID regularly with my yacht auto pilot (which there has been a few posts about) which is a quick changing PD and work with process PID's which change relatively slowly there for maybe easier to tune, but like Peter and now yourself (just a little bit) and l, l just can't believe how poorly the op's approach their projects/problems. I know a while ago a long time member played up (said he wasn't coming here again) l believe because of a post l wrote, that the person should read the manual, he took offence of what l wrote, it seemed he felt that any idiot should be able to ask a question here (which they can) and it seemed that he felt that people should pass on very hard earn't experience learnt over count less decades, because said idiot asked a question without even opening the manual.
Next ***** is someone who starts the thread with "URGENT HELP REQUIRED" someone helps and the op never comes back to say thanks or explain what fixed the issue for others to gain possible answers for later.
I personal can't understand why the specialists on this site even answer any questions, l come from the school of work it out for yourselves, if you can't work it out, you shouldn't be in the position you are in.
Like this thread, the op's should read the answers given and answer the question asked, so people here don't just go round in circles, how hard can it be.
Peters image is a classic, someone pulling tooth along. mark.t should answer questions, or not ask and go buy a tuning software for $1000's, or read some books, THEN come and ask more pertinent questions and supply answers to questions asked.
Atleast he didn't ask what is a PID, because l have to tune one:ROFLMAO:
 
...
Peters image is a classic, someone pulling tooth along. ...


:ROFLMAO::ROFLMAO::ROFLMAO::ROFLMAO::ROFLMAO:

Thanks 👨🏻‍🏫! I can be pretty t'ick-headed at times regarding jokes*, and didn't get Peter's until now.


Anyway, thanks for your thoughtful post. I think, given the nature of the industry, many OPs are in over their head as well as a desperate time crunch, and posting here is one of several Hail Mary's that they forget about. So maybe we can cut them some slack. Maybe.



My pet peeve is threads that ask about scaling, because, I'm sorry, they are just in over their head: proportions** were covered in grade school for heaven's sake!



* As a child it took me over a decade to get "How to you get down off an elephant?"
** 90% of engineering
 
My pet peeve is threads that ask about scaling, because, I'm sorry, they are just in over their head: proportions** were covered in grade school for heaven's sake!
Ditto. I see a lot posts and where the OP is clueless and shouldn't find another job.



I also think the OP's managers should be fired. It is obvious they are just as clueless as the OP. It is obvious they threw the OP under the bus on this one.
 

Similar Topics

I have S7 1512C controler for controlling 48 PID temperature loop, the output is PWM. Please I need the best, most efficient way to write the...
Replies
5
Views
218
Hi all, I'm having trouble solving a problem I've been working on for several months, and thought you might like a stab at it. The machine runs...
Replies
22
Views
851
How can I connect PID Output to a valve. In ladder logic program is there any logic do I want to add between valve and PID? PV=SP What will be the...
Replies
7
Views
385
I'm setting up a PID for a 50hp RO pump in a water treatment plant. When I go to manual mode on hmi I set the swm bit, I also ons the pid output%...
Replies
5
Views
524
I am setting up control for Hypochlorite dosing. The easy part is the dosing calculation for flow pacing but I would also like to setup trimming...
Replies
8
Views
895
Back
Top Bottom