For discussion of all aspects of measurement and control in the process industries, such as pharmaceutical, power, pulp and paper, chemical, etc.
If you have any questions or comments, please post them here. If you see a question that you can answer, or perhaps add your insight, please use the reply to add to the thread.
IEPAS02 power module INFI 90 Elsag Bailey
by
I'm looking for electrical drawing of the IEPAS02 power module INFI 90 Elsag Bailey
Thanks
sir
i want to know that when immerse conductivity probe in stirred tank
even with ordinary water it's showing fluctuations in reading where as if putoff stirrer then it's showing constant value ,iwant to know why it is showing so,my stirred tank is baffled so that no air is drawn
can you help in getting c-code for self tuning PID controller
by
please help me to get C-code for self-tuning PID controller in which i can place my desired transfer function and obtain desired out put for various set points
help in getting c-code and matlabcode for water level control
by
sir,
i am studying M.E control and instrumentation.
i am doing my project in conical water level control.
thus, i need c-code and matlab-code for this project.
so pls kindly help me.
thanking you
subha
i m also doing ME same branch as of u.my projet is "Liquid level control in conical tank ".my project review is on next week.i m in difficult in finding the mathematical model of it.
pls help me what r the references u referred.i will we kindful if u respond.
i m eagerly waiting for ur reply.pls mail to " jsarav78@rediffmail.com "
Dear friend,
Don't worry. God will help you. Search in the google by typing these words, " Matlab code on intelligent controllers". You can get some similar topics relating to your project.If you have any doubts please send me reply.
With regards,
S.Janarthanan
hi shilpa
me too like u am doing proj on level process station.
i found out that u asked for a code in this site.
so if u hav got this code,plz do also help me out.
kindly hope so
thanks
vijay.
can you help in getting c-code for self tuning PID controller
by
please help me to get C-code for self-tuning PID controller in which i can place my desired transfer function and obtain desired out put for various set points
Hi.
Now I've read some papers about cascade
control. And how I understood it can
help in preventing temperature lag.
It is very usefull for radiator control.
But I am interesting in cooperative
working of several controlled devices:
1.radiator(heat exchanger);
2.warm floor;
3.fancoil;
4.ventilation.
I think they cann't work simultaneously.
But how?
Maybe use some rules?
Maybe from value of derivative of temperature?
When big turn on radiator in other case fancoil.
Any ideas?
Sorry,Control Novice.
I thought that answer will
be sent to my e-mail. And then I
didn't mentioned the answer.
But now can You advice me good
literature, links for this subject.
Or maybe similar projects.
Thanks a lot.
Sorry again.
I have to design control system for
temperature control in the room.
There are several heating elements:
1.heat exchanger;
2.warm floor;
3.fancoil;
Also there is ventilation.
Can anybody advice in which way I have
to move.
Maybe cascade control?
RE: help me on project"control of dc motor woth fuzzy logic"
by
What are you trying to control and under what circumstances? I assume this is speed control. Do you need to manipulate the speed or hold a constant speed while the load is changing?
Hai guys,
I am new to this site.I am doing my B.Tech in Electronics and Instrumentation Engg.(final year).At this juncture, I am in need to do project. I have refered to google.com to have any useful sites,but Iam unable to get it,and the projects are also costly. So please give me any tips to my project(in any branch of instrumentation engg.)that costs cheep.
Hi there,
I`m new to this group.I`m making my resume and need to put a PID controller based project in it.I searched on google but I`m getting relatively complicated projects.I need simple projects so that I can handle in job interview as well.My humble request to u guys ,please give me some webaddress.Also any good websites to know PID from scratch.I learned it a while ago.
Hi guys.
I am new to control area.
I have got a task.
There is room with central heating, fancoil, ventilation,
warm floor, one window, one door.
And I have to develop system that would be able
to keep steady temperature. I can control valves on radiator of central heating, on fancoil.
Also I have sensors inside room, on tubes, and outside.
Can anybody advise me some way of solving.
I see two ways:
1. Use PID without building full model.But
what to do with persistence of heating elements
2. To build linear model but I don't know how.
I'll be appreciate for any idea or link.
Thanks.
I think a simple solution is cascade control. Read about cascade control at www.jashaw.com.
The secondary loop would have a temperature controller using a temperature element directly in the exhaust of the furnace. This may have a setpoint of, say 115 deg F.
The master loop would be the room temperature controller, set at, say 68 deg F. This controller would send the setpoint to the secondary loop.
So, if the outside air temperature changes, the secondary loop would see this affect quicker and adjust the radiator valve as necessary to maintain the master loop setpoint.
X which stabilises A+W*X(real parts of eigen values are <0) ,it’s existence can
Found out by
Eigen values of Hamiltonian matrix H given by
H MATRIX=
!A W!
!Q -A!
because we have the relation
EIGEN VALUE OF H ARE GIVEN BY= EIGENVALUES OF (A+W*x)& - (A+W*x);
In text it is stated as if there is no eigen values of H are on imaginary axis then X exists
Means it can have in real parts of ( eigen values can be >0)
This can be possible
If A+W*x has negative real parts
And also A+W*x has positive real parts in which it is un stable
If it is so how can we say that just H matrix not having eigen values on imaginary axis is
Sufficient for X toexist
Can any one explain me about this
Thanking you
I'm looking to implement Iterative Feedback tuning (IFT) in the tuning of a PID controller for a project I am working on for my final year BE. This is a technique that looks very interesting and uses no model so is very beneficial for our application. However, I find most material on it presents a bunch of abstract math and is very difficult to understand or visualise how indeed it could be implemented in practice.
If anyone has any information on this topic, especially in regards to finding the updated controller I would be extremely grateful.
hai andrew, i saw your mail in process control and measurement,today itself. iam also doing my B.E project in iterative feedback tuning of pid parameters. if you have any information about its implementation in any programming language, please send to me.
thanking you,
yours friendly,
pavan
hai andrew, i saw your mail in process control and measurement,today itself. iam also doing my B.E project in iterative feedback tuning of pid parameters. if you have any information about its implementation in any programming language, please send to me.
thanking you,
yours friendly,
pavan
hai andrew, i saw your mail in process control and measurement,today itself. iam also doing my B.E project in iterative feedback tuning of pid parameters. if you have any information about its implementation in any programming language, please send to me.
thanking you,
yours friendly,
pavan
hai andrew, i saw your mail in process control and measurement,today itself. iam also doing my B.E project in iterative feedback tuning of pid parameters. if you have any information about its implementation in any programming language, please send to me.
thanking you,
yours friendly,
pavan
The gain is the ratio of a change in the output to the change in the input that caused the output change. It can go from 0 to as high as you want (if you make it too high, the process will oscillate). Gains are very often between .5 and 3, but can be almost any number.
Reset rate is the repeats/minute. (the tutorial on the site will provide a better explaination). The code on the site is designed to be executed every second, or 60 times per minute. I don't think the reset rate should be above about 30 or it will not respond well. You can execute the code twice per second and change the constants that are now 60 to 120, and increase the maximum reset rate to 60. But I don't think it ever needs to be above 30. The reset rate can be as low as zero.
The derivative is in minutes, is typically less than 2, but can be anything from 0 to about 30.
Go to the Rosemount website. There are two, the gas division and the liquids division. You can get there by using the pull down menu on the upper right corner of www.emersonprocess.com and look around in documentation section. Excellent articles.
I am a forth year student of electrical engineering and my final design project is to design a small robotic arm to interact and play noughts and crosses. I need to be able to control four stepper motors 5V 4 pole each.
I need the code and software to complete this project could any body please help me it would be great.
When using a reverse acting flow controller which gives a pneumatic output, is it necessary to use a reverse acting valve positioner on the supplied control valves or can a direct acting valve positioner be used?
Any info on this or usefull sites would be appreciated.
As long as the controller is set up so that if the flow increases, the output decreases (reverse acting controller) and the valve is air to open, the positioner should be direct acting. An increase in signal to the positioner will result in an increase in valve opening.
Normally you will use a direct acting valve positioner. It has nothing to do with your controller except for the fact that you may have to change you controller action depending on which way your valve positioner is set up. The only time you want to use a reverse acting valve positioner is when you want to "reverse" the fail action of the valve. (Not meaning loss of drive air, more meaning loss of signal)
Example:
A flow controller and a fail closed valve. The positioner is direct acting. An increase in controller output opens the valve. Therefore you want a reverse acting flow controller to have negative feedback.
Now the same flow controller and fail closed valve. This time the positioner is reverse acting. Now no controller output means full positioner output and the valve is open. Full controller output means no positioner output and the valve is closed. Because the valve acts the opposite way (essentially, the controller thinks you have a Fail Open Valve) you now need a direct acting controller. Draw it out if reading it doesn't make perfect sense.
A new polynomial control method: Coefficient Diagram Method
by Serdar Ethem Hamamci
The Coefficient Diagram Method (CDM), recently developed and
introduced by Prof. Manabe (http://www.cityfujisawa.ne.jp/~manabes/)
in 1991. CDM is an algebraic approach applied to polynomial loop in
the parameter space, where a special diagram called coefficient
diagram is used as the vehicle to carry the necessary information, and
as the criteria of good design (Manabe,1998). The performance of the
closed loop system is monitored on coefficient diagram. The simplicity
of the controller structure made it very powerful for systems with
uncertainties such as robotic manipulator
(Hamamci(http://web.inonu.edu.tr/~shamamci/) and Ucar, 2002).
Generally, the problem of a control system design consists of choosing
a proper controller considering the system dynamics, which is to be
controlled, and desired performance specifications. There are three
main theory for a design procedure: Conventional Control Theory,
Modern Control Theory and Algebraic Approach. The main difference
among these theories is the design approach used to obtain the
controller and the mathematical expressions used to represent the
system . Classical Control Methods, such as Frequency Response Method
and Root-Locus Method, use the transfer function for the system
representation. However, this representation can lead to undesired
results because of pole-zero cancellations due to uncontrollable or
unobservable situations. Modern Control Methods, like Pole-Placement,
Optimal Control (LQR) and Hinf , use state-space representation. This
representation, especially as the plant degree gets larger, involves
complex calculations which require the use of a computer. Algebraic
methods like Pole-placement direct method and CDM use polynomial
expressions. In this representation, since the numerator and
denominator of the transfer function are considered independently from
each other, better results can be achieved against pole-zero
cancellations. In this approach, the type and degree of the controller
polynomials and characteristic polynomial of the closed-loop system
are defined at the beginning. Considering the design specifications,
coefficients of the polynomials are found later in the design
procedure. In algebraic methods, CDM is the one which gives the most
proper results with the the easiest procedure. In CDM, design
specifications are equivalent time constant , stability indices and
stability limits. These parameters have certain relations which will
be explained later with the controller polynomials (Manabe and Kim,
2000).
REFERENCES
1. S. Manabe, "Coefficient Diagram Method", 14th IFAC Symp. on
Automatic Control in Aerospace, Seoul,1998.
2. Hamamci S.E, Ucar A, "A Model Based Control for Uncertain Systems",
Transaction Institute of Measurement and Control, v.24, no.4,
pp.331-345, 2002.
3. S. Manabe and Y.C. Kim, "Recent development of Coefficient Diagram
Method", ASSC'2000 3rd Asian Control Conference, Shanghai, 2000.
Dr. Serdar Ethem HAMAMCI
----------------------------------------
Inonu University
Engineering Faculty
Electrical&Electronics Eng. Dept.
44280 Malatya/TURKEY
e-mail: shamamci@inonu.edu.tr
This is not a well established method. I have seen the papers of Hamamici and I do not think its worth publishing results of your kind. This method is hard to work for higher order processes .....especially if you intend to design a PID controller.
Is this homework for a class? If so, be sure to cite this forum with your answer.
We need to know the manual reset, or the value of the output when the input equals the setpoint. Otherwise, I will have to assume it is 100%, not a good assumption.
Setpoint conversion to percent:
Input range 5-20 bar, span is 15 bar.
11 bar - 5 bar = 6 bar. 6/15 = 40%
The output is 100% when the setpoint equals the measurement.
(15ma - 4ma)/16ma = 68.45% output.
The output will have to fall from 100% to 68.45%, or 31.25%.
A proportional band of 40% means that an input increase of 40% will result in an output decrease of 100%.
We need an output decrease of 31.25%, so we need an input increase of 31.25% x 40% = 12.5%.
The input span is 15bar. 12.5% x 15 bar + 5 bar = 3.25 bar.
So the answer is 3.25 bar if the manual reset is 100%
I have a motorized valve which controls the gas to a furnace. When I put 24V on one terminal it starts to open. When I remove the 24V the valve stays where it was. When I put 24V on the other terminal it starts to close. It has limit switches at the 'maximim open' and 'maximum closed' positions to prevent overtravel. It does not have any position feedback signal output.
How do I connect this valve to a PID temperature controller that has one 4-20mA analog output (I am not sure what the voltage is at the output)?
You want a PID control with digital outputs (i.e. output one=valve open, output two=valve close)Even one output would suffice, as you could control the valve with a relay (1 n/o contact, 1 n/c contact).
Gentlemen !
I see there are many clever brains out there and therefore I am raising my plea to you for a control solution to our "plant":
The plant is a so-called Rolling Road (=chassis dynamometer). This is a means of performance testing a road car: the driving wheels are running in a roller set, which is usually embedded in the ground. This roller set can be "braked" using a so-called Eddy current brake (=retarder), the type which is also used on heavy trucks/tractor units.
I am struggling to get a really good controller working for this.
The requirements are:
- minimal overshoot or no overshoot
- fast response
The roller set needs to be governed to run at a constant speed.
A speed signal 0-5V is avialable.
A load signal (how much the retarder is "braking" is also avialable as 0-5V).
As controller I have available:
MCU PIC 16F877
analog inputs to set GAIN, INTEGRAL, DERIVATIVE
analog inputs to sense speed, load
analog input to sense SETPOINT
I program the PIC in C (as I am not clever enough for Assembler :( )
I would like to try different control algortithms to fulfill my control needs. The person who priovides me with a solution which will find to work to my requirements I shall award 50 (fifty) US$. This I promise by my honour. This is only for solutions which work well !! For a poor solution I have already (my own :( ...).
If you would kindly send me some algortihms (Basic would be fine) or C for me to try, that would be great.
My e-mail is acc002@yahoo.com . It will take me a while to try them out, but I shall keep you posted on this interesting project.
This Rolling Road is for the means of performance tuning engines and we'll have Porsche , BMW, etc. on it. It will be great :)
It is important that the control ensures that the roller speed does not drift off, when the load (=braking) is varied. The roller speed must be kept constant as best as possible.
Maybe a "Fuzzy" would be better ?
Whatever, I will try all possibilities using the PIC 16F877 MCU (8 analog inputs, 20MHz clock). The MCU control output is an analog signal, by the way (digital outputs load a DAC with the output control signal).
Gentlemen, many thanks for your valued help.
Ladies, also invited, of course !
Many thanks !! :) :))
Rudi
P.S. example: if the setpoint was to be 50mph, I would like the rollers to run freely with no braking on at all until maybe 45mph, then the brake should be activated so that there will only be a minimal overshoot of say 10percent (=55mph) and a fast settling. Also, a "lerning" controller..Fuzzy ? would be a good idea..what do you think ?
PID control for AC step motor with big period of measure
by Artem
Hi.
I'll try to control current in phase of step motor. What relation between Kp Ki Kd I should make for good control,if I use PWM and can measure current one time per 4 periods of PWM?
sir,
I am trying to write a MATLAB code on "velocity form of PID controller" but not able to do so as it requires some past data.
can u please refer me the code.
--eagerly looking forward for ur reply.
--thanking u
--regards
-Rajesh Shaw
sir,
i am doing project on 8051 microcontroller based pid temperature control.so i need pid control algorithm in 8051 assembelly language. please send me immediatley.
thank you,
prinso.
sir,
i am doing project on 8051 microcontroller based pid temperature control.so i need pid control algorithm in 8051 assembelly language. please send me immediatley.
thank you,
sslonkar.
My PLC manufacturer (A-B) made a subtle but important change in the latest firmware release. I'm having to read between the lines in their docs, but it has to do with clipping the proportional term in the PID equation based on output limits. If you've specified the typical 0-100% limit on your controller output, and if the proportional term was driving the output beyond those limits, the old version would clip the proportional term. So if you had a large Kp and the operator made a large setpoint change, and the proportional term calculated an output of 800%, it would clip the proportional term to respect the 0-100% limit. The 800% would be clipped to whatever made the final output 100%, based on the contribution of the other terms.
The new version only clips the final output of the algorithm and retains the full value for the proportional term. This means keeping the full 800% in the algorith, the net effect being that the output will remain much higher as the process variable approaches the setpoint. The likelihood of overshoot is much increased, and retuning the loop is almost certainly required. We're either going to have to use derivative control or switch the proportional control to PV-based rather than error-based so that the output comes down sooner as the PV approaches the setpoint.
Is there a correct or standard way of handling output limiting? A-B claims that their new way is more correct. How do other manufacturers handle this?
I would like to see some literature from A-B about their PID equation. In general, PLC vendors have not had a record of understanding the complexities of PID (and related) control, compared to the major DCS vendors (Taylor, Foxboro, Honeywell, etc.)
This is from Peter Nachtwey by e-mail:
The new implementation of PID is more correct. All the terms should be combined first before limiting. This permits the digital system to more closely act like a analog controller. The problem with limiting only the output is that it is possible for the terms to overflow before limiting. Using floating point or 32 bit integers makes this unlikely but 32 bit PLCs and floating point requires more resourses.
I think you're correct. My problem is that their function block doesn't let you see the individual contributions and doesn't set a bit if the proportional contribution is greater than 100%. The operators, and I, get a little confused when we see a process variable above a setpoint and the controller output steady at 100%, not coming down as it should. I'm suggesting to AB that they set a status bit so you know this is happening.
I'm still curious how other vendors handle this, though. I haven't found sufficient detail in on-line docs from other vendors to indicate how they handle clipping.
Something seems very wrong here. I would say that there is a serious bug in their software.
I have seen implementations of PID where (assuming reverse action, output at upper limit) the output comes down from the limit before the process reaches the setpoint (depending on the rate of change of the process, and implementations where the output stays at the limit until the exact time that the process crosses the setpoint. I cannot imagine a controller where the output would stay at the limit after the process crossed the setpoint.
I would contact AB immediately and have them determine where the problem is.
Having worked for a major vendor (Taylor Instruments) for many years, I know that any new software release contains bugs, and the bugs will be fixed only if they are reported.
In this example, the AB function block will lower the output ... if you wait long enough. But if you're carrying an unclipped proportional contribution that is 200%, you won't see the actual output drop until the integral term has knocked out the excess.
Our problem is with a batch heating operation. If the operator has the setpoint at 75F, loads some partially pre-heated flavor into the tank that is 120F, then puts the controller in auto and punches in a 150F setpoint, he gets a proportional increase based on the change in error of 75F from his setpoint change. When the tank reaches setpoint, he has gotten back only 30F worth of proportional action from the gradual reduction in error from the PV rise. That leaves him with an excess proportional term that will only disappear from the contributions caused by crossing the setpoint. If the proportional gain was sufficiently high, this excess will hold the output at 100% for some time after the PV exceeds the SP.
AB says this is correct, and it does follow the algorithm that they describe. What it means is that you'd better use "SP-tracks-PV-in-manual" to avoid the above scenario. You may also want to consider making the proportional term based on PV rather than error so that setpoint changes don't generate the proportional kick in the first place. (But that's one reason why you use proportional control in the first place ... to get an immediate response based on a setpoint change.)
AB's software doesn't have a bug, as best I can tell. It's just that not clipping the proportional term can have some odd side-effects in batch heating operations like this. That's why I wondered how other manufacturers handled output limiting.
I work with the Bailey Infi 90 system and I work with their PID controllers. I'm not going to profess to be a know-all here, but I can tell you a little. In their PID block (also called Function Code 156) there are some user-defined specs for upper and lower limit of the controller output. When you look at the detail of the controller, this is not just on the final output. These parameters are also used on the integral term before it is added to the proportional (in a classical, interacting algorithm. With this controller you can choose from 4 algorithms). There is also a spec to determine how the controller responds to saturating the output due to integral. It's too much to write here, but it can be selected for quick or conventional saturation recovery limiting.
I don't know what AB has now for PID, (in their ControlLogix stuff) but if I remember right their PLC 5 PID is not as comprehensive or as flexible as Infi 90. But then a PLC 5 wasn't built for analog control either as a DCS is. It's not comparing apples to apples, as I'm sure you're well aware of.
Thanks for the Bailey info. While the PLC5 was limited, the ControlLogix implementation is much more full-featured. AB partnered with Honeywell on portions of PlantScape and picked up some process knowledge from them. I'm from a DCS background and have been quite pleased with AB's ControlLogix PIDE function block, at least up to this current issue.
AB does have features to handle reset windup, which I think are pretty common with all the major vendors these days. Not so sure about "proportional windup", which is what I'm calling this problem with unclipped proportional contributions to the final output.
That still doesn't sound right. Once the PV reaches the SP, the error is zero. Once the PV crosses the SP, by even the smallest amount, the error has changed signs and the output should start pulling away from its limit.
I still think it is a bug. Having worked for a vendor (ABB, now the owner of Bailey) I can say that the standard response is "that's the way its supposed to work".
Does anybody know a good method for doing a gradient smoothing on a grid of data points?
I would like to set hard limits on the gradients or use some kind of nonlinear function for it. There is a strong limit on computational effort.
If the routine would be in a recursive way, it would have to be limited to a certain recursion level.
The best approach would probably only involve changes neighbouring points and checks on their neighbours.
The data grid would have 400 gridpoints in a 20x20 grid.
Both gradient and absolute limits are set with a single value calibration that is used for all gridpoints.
It will be used on a fast DSP adaptive precontrol solution.
Please let me know in present situation how we can differentiate a PLC or DCS OR what are the differences between PLC & DCS, based on their functionalities, size etc.
Has a lot of usefull information about standards and technologies used in Process Control and Industrial Automation - Control Systems, HMI, Batch Control, Fieldbus, Advanced Control, Instrumentation, OPC, ...
where Error(t) is equal to Setpoint - Process_Variable at time 't'
I am trying to figure out the recursive equivalent equation of the above equation. I did some rough estimation and I'm not sure if my method is correct.
(i think!)The recursive form for the above PI control law is;
I am interested in being able to calculate the output (t) of the PI controller manually using a simple calculator to enhance my understanding on a simple PI controller.
Pls help.
Thanks.
If there are additional information that will aid in my understanding,pls email me triggerfish76@yahoo.com
The K_propPcnt/1000 is in the place in the equation that should be gain. Some controllers us gain, which is the ratio of the output change to the input change. The high the gain the faster the controller reacts. Gain is usually a dimensionless number, a gain of 1 means that a given percent change in the input results in the same percentage change in the output.
Calling it propPcnt makes me worry that the controller is tuned in proportional band (PB) PB is 100/gain, and is measured in percent. A PB of 100% is the same as a gain of 1, PB of 200% is the same as a gain of 0.5, etc.
Reset or integral may be in the form of Reset Rate in repeats per minute or reset time in minutes. If you increase the reset rate or decrease the reset time you speed up the control action.
The deta_t/T_reset is apparently the reset time in some units of time (minutes or seconds) and the deta_t is the amount of time (minutes or seconds) between scans of the algorithm.
Re: Enquires on Positional + Velocity Type of PID output
by Emiel
I think the author meant that he can control a velocity or position directly through changes on the output of the controller.
For example if you apply a voltage to a DC motor it will determine a speed, as speed is proportional to the applied voltage. If the controller would control current then this would have a proportional influence on the delivered torque.
The kind of consequence an action by the pi controller output has, is purely dependent of the connected process and the inputs and measurable outputs that are available. On this point one should always consider to design or find those inputs and outputs that make the process easier to control. This is most of the time a pure matter of the dynamics that is behind it and this involves either a frequency/phase spectrum analysis or practical trial and error (as is most frequently done for PI control)
I don’t often hear the terms "Positional" or "Velocity" (or the equivalent,
”full value” and “incremental”) anymore.
Back (1970’s through 1980’s), control was often implemented on a mini-computer connected to analog instruments on a panel-board. The analog instrument would provide the 4 to 20 ma signal that actually was connected to the equipment in the plant (typically a valve). There were two ways to connect the output signal to the analog instrument using a Digital to Analog converter in the computer.
One method, the velocity or incremental method, transmitted a rate of change to the analog instrument, which then integrated the rate of change to provide the actual current output to the valve.
The other method, the position or full value method, transmitted an analog voltage that corresponded to the actual position of the valve.
The advantage of the velocity form is that if the computer failed (quite common in those days) the analog instrument held its value and allowed the operator to adjust the output manually.
The advantage of the positional form is that the compute actually determines the valve position, can indicate it to the operator, and allow the operator to set the valve position through the computer.
The PID algorithm had to be different for the two methods. The velocity form will be integrated in the external analog device, so the equation is actually the derivative of the PID algorithm:
DeltaOut = Kp(dE/dt + Ki x E + Kd x d2E/dt2)
Where Kp is gain, Ki is Reset Rate, Kd is Derivative, E is error, and DeltaOut is what is actually output to the analog device.
ELast is the error from the previous execution of the code, ELastLast is the error from two passes before. The code is executed at some rate, perhaps once per second.
For the position form, the above equation is integrated. So the code would be:
The main difference in the two forms is how the integrator behaves. The position or absolute form can not unwind the integrator without over shooting the setpoint. The velocity or incremental form does not have a separate integrator and the P and D terms can 'unwind' the output eliminating overshoot if properly tuned..
I'm going to compare a fuzzy logic controller(FLC) with a conventional PID controller and i need to matematical and dynamic model of hydronic heating systems.
iwant to learn about flc and i want to send my the document about this and iwant to help in my project in msc to control the water level tank in fuzzy control to keep the level constant .please i want the program used in fuzzy control used in real time to control the data that come from isa bus
I am interested in determining the best methods of monitoring or evaluating controller performance. I am familar with the use of performance indices such as IE, IAE, ITAE, ISE, and ITSE in order to evaluate the deterministic performance of the controller. I have also come across the use of MVC (minimum variance control) when evaluating the stochastic performance of the controller. I am interested in finding out about any other techniques available that can be useful.
The methods such as ISE, IAE, etc. are useful in control studies on a simulator; for exsample, to compare one algorithm to another or one tuning method to another. They are not as useful on real processes where the difference in results from one time period to another may be unrelated to the controller. (amount of disburance, etc.)
To evaluate a controller's performance you must first define what is important. For example:
deviation from setpoint, by any amount,
deviation from setpoint my a given amount,
steady controlled variable even if not at setpoint,
response to setpoint change,
excursions from setpoint due to large disturbances, etc.
With the criteria based on what is important to your process you can then develop means of monitoring the controller performance.
I use ISE for most of the time for evaluting the tuning of motion systems. Motion system have a motion or target generator that generates a smooth transistion of setpoint from one point to another. A motion system should never have a large error. A good motion system does not make set jumps in the set point.
It seems that process system make step jumps in the set point. In this case the ITAE would be better as it would do a better job of evaluating settling time.
I haven't heard of MVC. I assume you are refering to linear quadratic control or something like that.
Your title mentions temperature and your message mentions flow. Are you asking about both?
For either, the transmitter converts the signal to a 4-20 ma. signal. The actual circuit depends upon the type of sensor and the manufacturer, but basically they have a circuit that ensures that the proper amount of current passes through the circuit depending upon the measured value. The sensor has a range, say 32 to 200 deg. f or 0 to 100 gpm. If the value is at the low end of the range, the transmitter will allow 4ma to pass through it. At the top of the range, the transmitter will allow 20 ma to pass through it.
In most cases the sensor and the transmitter are one unit. If you speak of a "flow transmitter" you are refering to the part of the device that connects to the pipe to sense the amount of flow and the electronics that control the current flow.
These days a growing number of transmitters send the signal using one of several digital standards, such as Foundation Fieldbus.
Anyone have schematic for a Eurotherm 832 power supply?
The one on Eurotherms web site (831/832) is not the same and there was another manual produced for the 832.
thank you
Doug
Almost all controllers on my chemical plant DCS are PI with no derivative. There are a few on temperature control that do have derivative. All, PI and PID, seem to be working well.
Is derivative appropriate for temperature control?
Yes, derivative is appropriate for temperature control.
Back in the days of analog control, you could order a PI controller or a PID controller for slightly more money. You had to choose up front, and justify even a small difference in cost.
Many engineering firms specified PID control for temperature loops and PI control for all other loops. With a DCS you can choose, at the time you are tuning the loop, whether or not to use derivative. However, the old rule of thumb: derivative on temperature, is still often followed.
There are several reasons for the old rule of thumb: derivative usually should be used on loops with very little noise. Temperature loops were usually very smooth and without noise. Derivative helps most when there are multiple lags in the process, again, typical of temperature loops.
Derivative can be used on other loops, and, with DCSs it does not add any cost or require a decision before the startup. You can try it. If it helps; use it; if it hurts, don't use it.
The short answer is yes.
The next question should be why.
Whether the derivative is in the forward ( using error ) or feedback ( using only PV ) part of the loop, it allows one to place another pole. This allows one to get a faster response. Time is money.
The next question is: why are you using only PI control?
There are some cases in motion control where imaginary poles, or lack of poles,make the D term almost useless.
Otherwise you should be using the Derivative.
Here's a different type of question for you
(maybe it's better on another discussion group though)
how do you explain basic control theory (at the PID level) to senior technical people who think they understand control theory but really don't?
it's a common problem in my life, it happens all the time. people just don't understand the basic way PV, Setpoint, and MV inter-relate. it's not even a calculus issue, they don't seem to understand the basic idea.
electrical engineers (as i am) generally do because i basically describe it as an ideal op-amp (do whatever you have to on the output to make the inputs equal...)
or maybe i'm just biased.
sounds cool, eh?
i've got to do this and i've got like 3 weeks to do it in.
the three jets are equally spaced around the perimeter of the device.
everything seems "under control" at this point of development, but i'm very concerned when i go to do this that the three loops will interact with each other (since they're coupled together by being attached to the same platform)
i.e. when a change height command is issued and the loops start responding, will the oscillate/fight each other?
will this be a problem? where do i find information/examples about doing something like this?
fun challenge, wish i had the time to enjoy the development process, but they're ready to ship the damn thing already........
Maybe a little late for an answer, but i think this is a typical problem where the influence will not be that big. You probably only use the controllers for height control at the three points. For a big hoovercraft (considering the thrusters to be in a triangle position at the corners of the vehicle) the height control could almost be seen as a pure vertical control of height for every separate thruster.
If you put the sensors for height measurement near the thrusters then the influence would be negligible and the height control could be done by three separate controllers.
This is not the case when you use four thrusters. Note the strong dependence of the diagonal corresponding thrusters here.
A last tip: Control is a matter of simple but good enough modelling. Often some simple force calculations reveal a lot.
Re: Re: Hoover Craft using 3 PID controlled Air Jets
by emiel
If the three fans share the same air container (what is normal for hoover crafts) then you would probably be controlling also a pressure target for the air under the vehicle. Of course this will act on all three of them. You could make a small model for decoupling the effects between the fans and use this for determination of the error that goes into the PID control.
i need pictures, plans and designs of a hoover craft, and detailed wrie up on how it functions all the dynamics , cos i have to build one for my project in 2 weeks so need all the information that can help me out.
steam pressure to temperature control with process controller
by rock
have a question that i hope someone can help me with.
have a old machine that is a autoclave machine. This machine has old antique controls that seems to be could be updated and simplified with a process controller such as a omron process controller.
Operation
1. Pull to -26psi vac inside vessel
2. Once reached , turn on steam valve, with pressure
transducer to measure pressure, once a certain pressure is reached, which equals temperature, then turn steam off and on to maintain pressure and temp.
Example - If temp inside vessel desired is 240 degrees then once the pressure in vessel equals approx. 10.96 psi - then temp equals 240F.
Could this be done with a process controller with the pressure transducer connected through a amplifier sending 4-20ma into process controller and use this to control output to 4-20ma contolled valve. Could someone please give advice and if possible, example of how to do this.
thanks
RE: steam pressure to temperature control with process controller
by
Yes, you can do this with the pressure input connected to a controller and the analog output connected to a control valve (usually a pneumatic valve using an I to P transducer).
It would be better if you could measure the temperature and use a cascade temperature controller manipulating the setpoint of the pressure controller.
use a pressure transducer instead of temp sensor, due to fact once autoclave chamber is vac to -26psi, and then steam turned on, steam is sucked into the chamber so quickly that a chamber 12ft in length by about 6ft in diameter will get to desired pressure of about 10.96 psi ( =240degrees ) very quickly and temp sensor not accurate due to the quick inrush of the steam.
would you have any ideas or suggest a controller.
thanks
Hello, I just wanted ot know how much of the Converged Technology blocks such as VoIP (voice over IP), VXML, SIP, SAP and the like are under developement in the Arabian Gulf. My expertise are in Telephony, Unix and Systems Administrations, want to know what or how far is the technological advance in the Arabian gulf, specially in Qatar and what are the hopes of this development in the next year or so?