wvreeven and I have communicated offline trying to figure out how to optimize the settings for the internal guider. We have not been able to glean all the necessary information from the description pages of the guide module: indilib.org/about/ekos/guide-module.html
We think that 'Proportional gain' is equivalent to 'Aggressiveness' in PhD2, but we are not certain about the effects of adjusting 'Integral gain'. Or are they equivalent to 'predictive gain' and 'reactive gain'.
Currently, our guiding accuracy in RA using the internal guider hovers between 1.3" and 2.5", we would love to make that more consistent and better.
It would be great to know if there are any detailed descriptions of the algorithm these parameters feed into, that would save us from reinventing the wheel by trial and error.
Anyone who has experience with adjusting these parameters and what they mean?
The most up-to-date documentation about guiding is here: stellarmate.com/support/ekos/guide.html
If someone has time to edit indilib.org Ekos documentation pages to sync them with StellarMate pages, please let me know.
Ekos guide code was based on lin_guider code from a couple of years ago. The algorithm acts like a
Proportional-Integral (PI) controller
. I'm not going to discuss control theory here, but the proportional gain (133.3x) is multiplied directly by the deflection value (+/- arcsecs) which is the difference between the actual star position minus the locked position. The integral term (sum of errors) is added to that, but the integral gain is zero is by default the integral term is not used. This produces the pulse that gets sent to the mount. So:
Pulse[ms]@T = deflection@T (arcsecs) * PropGain + sum(deflections from 0 up to T)/50 * IntegralGain
If you turn on debug logging, you'll see a LOT of output of all the math involved.
Excellent, thanks Jasem. That helps a lot with understanding how the guide pulses are calculated. We had pretty much correctly deduced from mount behavior how the proportional gain is calculated. From the definition of the integral gain it looks like the most appropriate course of action is to leave it set to 0 for mounts that have already undergone permanent periodic error correction or that fluctuate around 0. By contrast, for mounts with a largely constant inherent tracking error (either + or -) integral gain would be a very useful parameter to play with. My iOptron SmartEQ has that problem so I will change the integral gain settings when using that one.
We'll experiment with the settings based on these algorithms and post our experience, if useful.
As for syncing the Ekos documentation with the StellarMate pages: Looks to me like Ekos covers most of what is described on the Stellarmate pages minus a few details that you just filled in, at least as far as the guide module description is concerned. I would suggest to simply include the appropriate stellarmate links on the Ekos documentation pages at the bottom with the note that 'Additional information may be available at http://stellarmate......"
Also, thanks for pointing out the debug logging.
Jo