PID tuning, any good reads?

[Chris Wilson]

I am slowly getting my head around PID control of engine control functions like boost control, idle control, and one day traction control. There are many that say even at the OE engine calibration level these MATLAB type strategies for working out an algorithm away from the vehicle are not really used, as at the end of the day trial and error testing give more accurate and stable results. Can anyone share any real world experiences from either the automotive, avionics or industrial control sectors please? Am I wasting my time reading stuff in the hope that I can get close without actually experimenting in real time on the vehicle or engine itself? I was half contemplating getting this:

http://www.pidtuning.net/

His free tutorials have helped me grasp the concepts a lot better than trying to remember calculus Thanks.

[Benkku]

Sometimes starting values, are difficult to predict, hence trial and error method might be best during the field operations - especially in situations when actuator is not following command in a 'linear manner' - pneumatic positioning systems for example. Also there are quite few PID algorithms, behaving 'somewhat differently'.
Tuning is quite easy, if and when there are measurements logged - actuator(s) actual movement against control value (pressure, voltage etc.) and time - is the most important. So one can see how the controller/algorithm behaves - does it overshoot, undershoot or what.

Some good reads, then...
http://en.wikipedia.org/wiki/PID_controller
http://www.jashaw.com/pid/tutorial/
http://www.cds.caltech.edu/~murray/cour ... om-ch6.pdf

[Chris Wilson]

Thanks for those links and the info, very good of you, will download and have a look tonight much appreciated cheers