PID loops

S

steve cordiner

I've been tuning loops for years and only started using protuner which you hook onto the loop, do the porcess upset and let it do it's thing. Wow, 100% results, everytime. Downside is cost, £10K, but if you can borrow one or have several loops to tune, it's worth it's weight in gold!
 
P

piyushvastani

I've just automated a 50 foot long rotary kiln for the drying of coal. Now I am deeply researching PID loops. I need to adjust a gas valve to maintain a set point temperature coming out of the dryer... I was wondering if anybody might be able to send me any PID setting that they know have worked and maybe a little about the application they were in. I'm thinking I need to have a large derivative term, small integral term and somewhere in the middle for the proportional. But I'd like to study some examples.

This is a siemens s5 155 olc

THANKS!
 
Mr. Lloyd,
I too am impressed with the teaching, theoretical and practical performance of IMC and I'd like to take you up on your offer for more of your information on the subject. I'd greatly appreciate studying anything you could send me
([email protected]) on the subject.

Thank you,
Keith C. House
USA
 
R

Rick Carrasco

Have a look at a product called BrainWave at "www.brainwave.com":http://www.brainwave.com . It has been succesfully applied to many kiln applications and it can handle the dead time typically associated with kilns. It integrates well with a plc. It is a software controller that uses model based, adaptive, predictive control technology.
 
dear
hi
please tell us whats fully function of PID, because in books and other literature we study but not fully understand about PID, we have some query about PID, following query are as under.
1.Normelly we set PID P=100%,I=0.75,D=0.25 these values are ok in feed forwarded control?
2.If we want fast control to final control element which value of PID we change our Final valve action is fast?
3.if we want slow action to final control valve which valve we set for slow actiion.
if you understand my query please write us, i shall be very thankful to you.
regards
rana
 
Derivative action is instigated by the rate of change of the process variable (PV) from the desired set point (SP). If the control loop has a fast response, such as a gas flow control loop, then the rate of change of the PV will be relatively high and hence corrective derivative action will be applied. Due to this fast response it is very likely that the control will become unstable due to inherent nature of derivative action.

Slow response systems (systems with significant lag and large total capacity), such as a temperature control loop, are more suited to derivative action where rates of change in PV are much slower.

I have had experience in tuning temperature contol loops on gas fired batch kilns where derivative action works well - due to the system lag/reponse times.
 
J

Jeremy Buchanan

Robert,

Please understand that no one can give you the tuning parameters you need unless you have an identical process somewhere. You can use closed-loop tuning, but that usually requires expensive software that tracks output changes and PV responses on-line. This is convenient and perhaps more accurate than open-loop tuning, but very expensive.

Open-loop tuning is named such because it involves manual mode operation of the loop and recording of CV step changes and subsequent PV responses. The Cohen-Coon tuning method (http://www.chem.mtu.edu/~tbco/cm416/cctune.html), presumably developed by Mr. Cohen & Mr. Coon, is a generalized algorithm to convert the real world process responses and time delays into the tuning parameters you seek. I have had very good resutls using this method for outlet temp control of gas-fired rotary dryers and coal tonnage rate control. The Cohen-Coon gain is usually pretty "hot" in my experience - I recommend using no more than 75% of the calculated gain as a starting point to avoid angry operators.

After you've calculated the gain, rate and reset by whatever means, you must be careful to understand that these numbers may not be those needed by the PLC algorithm. For instance, a PLC-5 allows the programmer to specify whether the gain and reset are dependent or independent. The choice effects the values to be entered. The units of reset also pose an opportunity for confusion: ie: are the PID block reset units repeats/minute or minutes/repeat. It behooves you to answer that question before hitting the "Test Edits" button.

Jeremy
 
One thing that you must be careful with is the impact of larger tuning parameter settings in a PID controller. It's often confusing (this is very true for the Allen-Bradley PID instruction) whether a larger number represents more control action or less control action.

The proportional term is usually called "proportional band" or "gain". These have reciprocal actions, and the typical relationship is PB=100/Gain. For PB (the unit is usually percent), an increased value represents less control action. For Gain, an increased value represents more control action.

The integral term is usually called "reset", "resets per minute", or "minutes per reset". The latter two are reciprocals, while the first could be either. For R/M, an increased value represents more control action. For M/R, an increased value represents less control action.

It's important to study your controller documentation carefully. The A-B PID instruction supposedly uses Gain and M/R, so big numbers represent increased proportional action and decreased integral action. However, because the SLC-5/03 PID instruction uses a dependent PID equation, you can't get much integral action if the Gain parameter is small.

To anyone out there who's coding PID: Please, Please, Please use Gain and Resets/Minute. It's so much less confusing when a larger number represents more control action, and zero represents no control action.

Jeff M. Goldsmith, PE ([email protected])
Ionics Incorporated
Watertown, MA, USA
 
Hi,
would it be possible for to get a copy of the program that you have written to simulate a PID controller. It would be highly appreciated.
Thanks

pribhagan(AT)netcourrier.com
 
Just wondering with all this talk on "derivative"
action in relation to temperature control, is it worth testing the measuring element to get some
idea of response? That is finding the time constant of the unit including the well. In some cases there may be a significant lag in response to real temperature change. This sensor response then be inverted using lead/lag action to give a closer representation of temperature.

[email protected]
 
I've just automated a 50 foot long rotary kiln for the drying of coal. Now I am deeply researching PID loops. I need to adjust a gas valve to maintain a set point temperature coming out of the dryer... I was wondering if anybody might be able to send me any PID setting that they know have worked and maybe a little about the application they were in. I'm thinking I need to have a large derivative term, small integral term and somewhere in the middle for the proportional. But I'd like to study some examples.

This is a SLC 5/03.

THANKS!
 
K

Kaushik Patel

Hi Jeff

I am very much intrested in your PID loop VB,
I do not have VB5

Thank you and best regards

Kaushik Patel
 
C
Ny point is, that what you, (and I) consider pretty darn good is still nowhere near what the published figures would indicate. And that is because electronics, as a class, are pretty darn reliable. But when you have two systems, using readily available components, automated assembly, commodity connectors and competant design, you are simply not going to see the vast differences in reliability that exist in the minds of the marketers. And I haven't seen a true mil spec design for years. And that's OK because it costs a lot and still doesn't change things as much as some folks would have you believe. And much of the commodity stuff is more reliable than mil equipment from the recent past because experience is far more valuable in fixing trouble areas than prediction is. I participated in the game because because a number was demanded. Whether anything actually met that prediction is largely moot because it's obsolete long before it's dead. So nobody knows, you could use any number you want.

Regards

cww
 
I think derivative act as an anticipatory control, its anticipate the direction of error, thats why using larger value will coz your control go wilder (but ur in control) or change faster.
 
I have tried so many methods to tune a PID, some based on experience and some based on theory, and I've found that Ziegler-Nichols rule is the best when ever a log is available. The idea is to set the Proportional to a high value almost double of expected range. and have a log with time.

the final parameters will be directly calculated from a simple formulas.

you can find the complete method at page 673, Modern Control Engineering, Ogata, ISBN: 0-13-261389-1

ask Jr [email protected]
 
A

Alireza Askari

I think you can use PID controller with autotuning function. In this method your controller changes and plays with PID parameter and achieves the best parameter depending on your system specification. Besides this, you can use self-tuning function to recalculate the PID parameter periodically if behavior of your system changes.

Thanks!
 
M

Manmeet Singh

Dear Friend,

Just doing the basic ZN tuning for the PLC won't do. What you need is auto tuning software. This software will cut down on the time of tuning but the model based control inbuilt will help to cut down on the losses and coupled PID paramters for the process.

Need to understand the behaviour of the control loop. Otherwise the loop causes losses in energy usage.

Regards,

Manmeet
 
Top