PID tuning - Is there an easy way??
- Thread starter Cryogen
- Start date
Ron Beaufort
Lifetime Supporting Member
Greetings Cryogen,
while we're waiting for Peter, you might want to take a look at the blue (pressure) curve in Figure #2 of this thread ... I'm hoping that Peter will agree that the response curve I've drawn there is a simple example of a simple "first order" system ...
basic idea: it looks just like the "charging curve" of the capacitors that you probably studied in basic electronics courses ...
you might want to compare the shape of that response with the green (level) curve in Figure #5 of the same thread ... notice that one just climbs straight up until the tank overflows ... I've heard people call that type of response an "integrator" system ...
Peter will be along soon and provide more detail - and possibly a correction or two ... but in the meantime, this should give you something to get started with ...
while we're waiting for Peter, you might want to take a look at the blue (pressure) curve in Figure #2 of this thread ... I'm hoping that Peter will agree that the response curve I've drawn there is a simple example of a simple "first order" system ...
basic idea: it looks just like the "charging curve" of the capacitors that you probably studied in basic electronics courses ...
you might want to compare the shape of that response with the green (level) curve in Figure #5 of the same thread ... notice that one just climbs straight up until the tank overflows ... I've heard people call that type of response an "integrator" system ...
Peter will be along soon and provide more detail - and possibly a correction or two ... but in the meantime, this should give you something to get started with ...
rsdoran
Lifetime Supporting Member
A first order system is one that is governed by a differential equation that consists of only a variable and its first derivative.
These may help:
http://www.facstaff.bucknell.edu/mastascu/eControlHTML/SysDyn/SysDyn1.html
You have to have Shockwave.
http://en.wikipedia.org/wiki/First-order_logic
It is basically P&I but as mentioned Peter will probably offer the best info.
These may help:
http://www.facstaff.bucknell.edu/mastascu/eControlHTML/SysDyn/SysDyn1.html
You have to have Shockwave.
http://en.wikipedia.org/wiki/First-order_logic
It is basically P&I but as mentioned Peter will probably offer the best info.
Peter Nachtwey
Member
Control is all about energy.
Control is about how energy is converted from one form to another and how quickly it is converted from one form to another.
A first order system has 1 energy storage device. In your case it is the head in the tank. In a RC circuit it is the capacitor. Every device that stores energy adds to the order of the system. For instance a mass and spring would have two energy storage devices. The spring can store energy when it is compressed or stretched in the form of potential energy. The mass store energy in the form of kinetic energy. When a mass oscillates on the end of a spring the energy is transfer back and forth between the spring and the mass. Friction ( damping ) causes the mass and spring to stop oscillation as it converst the energy in the mass and spring to heat. ( Remember this dicussion? ).
As I said above, your case is similar to a RC circuit where the energy appied is not immediately seen at the output. It takes time to charge the capacitor just as it takes time to fill your tank. Your system has only one storage device, the tank. That is why it is first order.
Each kind of system will require different techniques for tuning. In your simple first order case I recommend that just do it because these are very easy to get close manually with a little trial and error.
Other systems may be 4th order systems. Sometimes the 'energry storage' devices can transfer energy back and forth like the mass and spring. In other cases they can't. In the case of two RC cicuits in series the energy goes one way. As I said above, each type of system requires a slighty different technique. A couple years ago I made some spread sheets to show how different systems are different in the way they tune.
Try this one:
http://www.deltacompsys.com/out/T0P1%20PID.zip
This is similar to what you will be tuning. After you get that one then try this one:
http://www.deltacompsys.com/out/T0P4%20PID%20ABS.zip
This one is MUCH hard because it is a forth order system. Pray you don't have to tune one because you really need more than just a simple PID.
Beware of those that provide tuning suggestions when they have no idea of what your system is like. I still will not tell you what Kc to use or what Ti should be. You will have to find that out on your own. I just know it will be a good system to learn on because it should be simple
Did the bias or feed forward make sense?
Edited additions.
Ron B. I don't like that example only because you used air. Air compressing or exanding is an example of a non-linear first order equation. The fact that this system is non-linear make the calcuations very difficult. I have examples showing air expanding ( escaping ) from a tank. Air is non-linear because it compresses differently as the pressure changes. The bulk modulus of air changes as pressure changes.
Ron, your examples are technically correct and are how I look at the problem. However, I am hoping that my non-mathematical energy storarge desciption is more understandable. How did I do?
Control is about how energy is converted from one form to another and how quickly it is converted from one form to another.
Cryogen said:
A first order system has 1 energy storage device. In your case it is the head in the tank. In a RC circuit it is the capacitor. Every device that stores energy adds to the order of the system. For instance a mass and spring would have two energy storage devices. The spring can store energy when it is compressed or stretched in the form of potential energy. The mass store energy in the form of kinetic energy. When a mass oscillates on the end of a spring the energy is transfer back and forth between the spring and the mass. Friction ( damping ) causes the mass and spring to stop oscillation as it converst the energy in the mass and spring to heat. ( Remember this dicussion? ).
As I said above, your case is similar to a RC circuit where the energy appied is not immediately seen at the output. It takes time to charge the capacitor just as it takes time to fill your tank. Your system has only one storage device, the tank. That is why it is first order.
Each kind of system will require different techniques for tuning. In your simple first order case I recommend that just do it because these are very easy to get close manually with a little trial and error.
Other systems may be 4th order systems. Sometimes the 'energry storage' devices can transfer energy back and forth like the mass and spring. In other cases they can't. In the case of two RC cicuits in series the energy goes one way. As I said above, each type of system requires a slighty different technique. A couple years ago I made some spread sheets to show how different systems are different in the way they tune.
Try this one:
http://www.deltacompsys.com/out/T0P1%20PID.zip
This is similar to what you will be tuning. After you get that one then try this one:
http://www.deltacompsys.com/out/T0P4%20PID%20ABS.zip
This one is MUCH hard because it is a forth order system. Pray you don't have to tune one because you really need more than just a simple PID.
Beware of those that provide tuning suggestions when they have no idea of what your system is like. I still will not tell you what Kc to use or what Ti should be. You will have to find that out on your own. I just know it will be a good system to learn on because it should be simple
Did the bias or feed forward make sense?
Edited additions.
Ron B. I don't like that example only because you used air. Air compressing or exanding is an example of a non-linear first order equation. The fact that this system is non-linear make the calcuations very difficult. I have examples showing air expanding ( escaping ) from a tank. Air is non-linear because it compresses differently as the pressure changes. The bulk modulus of air changes as pressure changes.
Ron, your examples are technically correct and are how I look at the problem. However, I am hoping that my non-mathematical energy storarge desciption is more understandable. How did I do?
Last edited:
rsdoran
Lifetime Supporting Member
Worked like a charm for me, actually wish I had thought to use something like that. That is the problem with knowing a little about something, you have a strong idea of the concept but not how to put it into words.
Peter Nachtwey said:
It has been my experience that on the SLC 500 the output has only 100 steps (Word 16: Output CV%(0-100%) ). Am I missing something?
EDIT ** EDIT ** EDIT
Never Mind. I found it. For some reason, I thought that the Control Output value would be in the PID control block.
Last edited:
Terry Woods
Member
- Join Date
- Apr 2002
- Posts
- 3,170
Step-size depends entirely upon the number of bits used to resolve the particular condition into a numerical value AND the full-range of the measured condition.
If based on a 0-MAXmA... (as opposed to 4-20mA)
If the AD Converter uses 3 bits... and the range is 0-100... the step-size is 14.285 units.
If the AD Converter uses 4 bits... and the range is 0-100... the step-size is 7.142 units.
If the range is greater... the step-size is larger.
If based on a 0-MAXmA... (as opposed to 4-20mA)
If the AD Converter uses 3 bits... and the range is 0-100... the step-size is 14.285 units.
If the AD Converter uses 4 bits... and the range is 0-100... the step-size is 7.142 units.
If the range is greater... the step-size is larger.
siva_maddala
Member
How To Tune Pid Parameters For Tank Control In Rslogix 500
Hai welcome to automation world
recently i am working on MICROLOGIX 1100 i did the operation for PID block. that is one tank is there and one inlet modulating valve is there so i have to control the valve using level of the tank. i created PID block and checked in manually it is working nicely whenever it is in auto suppose if the setpoint of tanklevel exceeds the level of the tank then valve is auto matically opening suppose then we reduced the setpoint to below the tanklevel then valve is not closing still it is opening, i set the parameters are
proportional gain 1
ti 1
td 0
E = SP-PV
TIMED
YES
DEAD BAND 25
SO IF ANY ONE KNOWS THE VALUES HOW WE HAVE TO SET IN PID BLOCK FOR MICROLOGIX IN RSLOGIX 500 Immediately send to my mail id. i am eagerly waiting for your reply
thanks and regards
M.SIVA
Hai welcome to automation world
recently i am working on MICROLOGIX 1100 i did the operation for PID block. that is one tank is there and one inlet modulating valve is there so i have to control the valve using level of the tank. i created PID block and checked in manually it is working nicely whenever it is in auto suppose if the setpoint of tanklevel exceeds the level of the tank then valve is auto matically opening suppose then we reduced the setpoint to below the tanklevel then valve is not closing still it is opening, i set the parameters are
proportional gain 1
ti 1
td 0
E = SP-PV
TIMED
YES
DEAD BAND 25
SO IF ANY ONE KNOWS THE VALUES HOW WE HAVE TO SET IN PID BLOCK FOR MICROLOGIX IN RSLOGIX 500 Immediately send to my mail id. i am eagerly waiting for your reply
thanks and regards
M.SIVA
Peter Nachtwey
Member
The problem with the simulation is that it is controlling the outflow and not the inflow as in siva_maddala's problem. Also, the simulation has a glaring error in that the out flow doesn't vary with the height of the the fluid column.
Mickey
Lifetime Supporting Member
Similar Topics
Hello, I am attempting to tune a PID loop on a process. The process involves a valve with electronic actuator that has quite a high deadband...
so i have 4 25gpm wells feeding a 1000gal tank (T-1), with an additional 15gpm from a decant tank for 3hrs every 12hrs.
P1 and P2 both controlled...
Hi everyone, yet another PID problem. I'm hoping I understand enough of the process I'm controlling that my request for help is reasonable. If a...
A few months ago, I started to look into PID controllers and the tuning of first order processes. This has, partly thanks to you, resulted in a...
I haven't had to tune a PID loop in a very long time.
It's actually a PI loop for a pulse width modulation s.v.
What was the name of that tuning...