Peter Nachtwey
Member
Re: Simplified output math
Excellent, variations/experiments are welcome. I agree that this is informative. One can see how the control output shift back and forth between the proportional and integrator terms. Finally the integrated gets wound up and it perfecty tracks the target speed and all the control effort comes from the integrator.
That is because the output does not saturate at plus or minus 10 volts. Try limiting the output to plus and minus 10 volts and let the target velocity do a step jump to 10 inches per second. You will see that the abolute form of PID will over shoot because the integrator will windup because the system is saturated.
All the time. Yesterday I modeled a hydraulic system that will be installed in a major tire manufacturer. There will be many systems so the object is to put a little more engineering effort in checking the design so money isn't spent where it didn't need to be. I do this because it makes every one look good.
If I didn't do this there is a chance that there could be a finger pointing contest.
Yes, but first you must have a model. Then you can CALCULATE the gains. Hopefully this is the starting and ending point. I can calculate the model if I have data the shows how the PV or actual position varies with the control signal. So far I have provided the model and equations for calculating the gains.
If you want a head start.
Later this summer I want to show how this can be done using RSLogix trend ( you would think the Rockwell guys would get me a copy of RS500 for doing this but noooooo, they charge me like everyone else ), and an Excel spreadsheet.
I have other things I want to cover first.
Tuning a 4 real pole system. What you have seen so far is easy. I haven't even calculated the formulas for computing the gains yet.
Tuning a system with imaginary poles. How to control a slinky. The classical weight on a spring.
When things are NOT pefect.
noise
quantizing errors
sample jitter
the affect of PID loop times.
dead bands
dead times
Different forms of PID
velocity mode ( incremental )
position mode ( absolute )
PID
PI-D
I-PD
etc. this might take more than the summer but I don't want to go to fast and I do want everyone to take the time to experiment with the spread sheets.
vern said:I created separate columns for the proportional, integral and derivative portions of the control output. It should help people who are not as mathematically inclined to understand a bit more.
Excellent, variations/experiments are welcome. I agree that this is informative. One can see how the control output shift back and forth between the proportional and integrator terms. Finally the integrated gets wound up and it perfecty tracks the target speed and all the control effort comes from the integrator.
vern said:I used Norm's idea for the shape of the velocity profile. You can see the separate effects of the P and I portions. The form for the PID output I used gives the same results as Peter's for this profile. I didn't put anything in to limit the output so it won't work for a real application. I really like the way Peter's form for the control output works to prevent the integral windup.
[/B]
That is because the output does not saturate at plus or minus 10 volts. Try limiting the output to plus and minus 10 volts and let the target velocity do a step jump to 10 inches per second. You will see that the abolute form of PID will over shoot because the integrator will windup because the system is saturated.
vern said:Thanks for starting this Peter. How much do you use system modelling?
[/B]
All the time. Yesterday I modeled a hydraulic system that will be installed in a major tire manufacturer. There will be many systems so the object is to put a little more engineering effort in checking the design so money isn't spent where it didn't need to be. I do this because it makes every one look good.
If I didn't do this there is a chance that there could be a finger pointing contest.
vern said:Do you use it to get a starting point for your gains?
[/B]
Yes, but first you must have a model. Then you can CALCULATE the gains. Hopefully this is the starting and ending point. I can calculate the model if I have data the shows how the PV or actual position varies with the control signal. So far I have provided the model and equations for calculating the gains.
If you want a head start.
Later this summer I want to show how this can be done using RSLogix trend ( you would think the Rockwell guys would get me a copy of RS500 for doing this but noooooo, they charge me like everyone else ), and an Excel spreadsheet.
I have other things I want to cover first.
Tuning a 4 real pole system. What you have seen so far is easy. I haven't even calculated the formulas for computing the gains yet.
Tuning a system with imaginary poles. How to control a slinky. The classical weight on a spring.
When things are NOT pefect.
noise
quantizing errors
sample jitter
the affect of PID loop times.
dead bands
dead times
Different forms of PID
velocity mode ( incremental )
position mode ( absolute )
PID
PI-D
I-PD
etc. this might take more than the summer but I don't want to go to fast and I do want everyone to take the time to experiment with the spread sheets.