Maybe someone is interested in this algorithm, which is a kind of PID controller but with better performances.
It is described in great detail in this document published by Zenodo. It is released under the Creative Commons Attribution Non Commercial Share Alike 4.0 International license, so it can not be used in commercial products.
Another explanation about D-PID is found in my site, https://linuxfan.altervista.org/pid/4-dpid/
The algorithm can be tested using a program written just to experiment and study PIDs, described here: https://linuxfan.altervista.org/pid/the-pidanalyzer-program/. The program is published on Codeberg, in my repository at https://codeberg.org/linuxfan/PidAnalyzer.
The thing boils down to this simple code:
Where:
PV is the process variable
SP is the set point
slope is a parameter, it controls the slope of the output curve
hard is another parameter, it controls the "force" of the PID
out is the output of the controller
I've successfully implemented this algorithm using integers (fixed point math) in a real-world application, and it worked well.
I hope this can be useful but I can not make any guarantees about the applicability of this idea, so I decline any responsibility.
It is described in great detail in this document published by Zenodo. It is released under the Creative Commons Attribution Non Commercial Share Alike 4.0 International license, so it can not be used in commercial products.
Another explanation about D-PID is found in my site, https://linuxfan.altervista.org/pid/4-dpid/
The algorithm can be tested using a program written just to experiment and study PIDs, described here: https://linuxfan.altervista.org/pid/the-pidanalyzer-program/. The program is published on Codeberg, in my repository at https://codeberg.org/linuxfan/PidAnalyzer.
The thing boils down to this simple code:
D-PID algorithm:
desslope = (sp - pv) * slope
realslope = pv - prevpv
prevpv = pv
out = (desslope - realslope) * hard + out
if out < -1000 then out = -1000
if out > 1000 then out = 1000Where:
PV is the process variable
SP is the set point
slope is a parameter, it controls the slope of the output curve
hard is another parameter, it controls the "force" of the PID
out is the output of the controller
I've successfully implemented this algorithm using integers (fixed point math) in a real-world application, and it worked well.
I hope this can be useful but I can not make any guarantees about the applicability of this idea, so I decline any responsibility.
