+- OpenMarine (https://forum.openmarine.net)
+-- Forum: Pypilot (https://forum.openmarine.net/forumdisplay.php?fid=17)
+--- Forum: General discussion (https://forum.openmarine.net/forumdisplay.php?fid=18)
+--- Thread: Three questions about gains (/showthread.php?tid=1996)
Three questions about gains - ironman - 2019-09-17
I'm in the gain tuning phase now, and I'm venturing into the uncertain realms beyond the easy-to-understand P and D gains.
First thing I have noticed that I cannot get ap.pilot.basic.Igain up. As a test, I've left a heading error build up for more than an hour now, but ap.heading_error_int stays zero and the 'bar' in the plugin gains screen does not go up. I'd expected it to be blue by now (see attached image). I checked the code and can't figure it out. What am I thinking wrong?
Second question is regarding ap.pilot.basic.R and ap.pilot.basic.Rtime. What are they and how can they be used?
Third question is about ap.pilot.basic.FF. Is this only useful with the autoroute plugin or NMEA APB (gps mode) heading changes, or does it also bump up reactiveness to manual (compass mode) heading changes?
Thx!
RE: Three questions about gains - ironman - 2019-09-18
Come to think of it, I would expect PRgain also to have a certain value, not zero.
Line 99 of basic.py in github says PR = math.sqrt(abs(gain_values['D'])); surely that must be 'P' instead of 'D', right?
RE: Three questions about gains - seandepagnier - 2019-09-20
1) I will look into
2) This is related to the reactive gain which is not very well tested/proven at this stage. If you set it to zero it should not have an effect.
3) The FF gain is the feed forward gain which steers when the heading command changes. This has effect when manually changing course while engaged, as well as autopilot route following.
4) I think you might have found a problem. Not sure how I overlooked this, but the PR gain did not work as intended as you say. I have changed it to use the position root rather than rate root.