by Hooray » Thu Apr 02, 2020 7:01 pm
yes, I've repeatedly done that before - and it works well enough, even without customizing the standard PID controller.
I don't see any reason why this should no longer work - the mode won't work as you expect, but you can definitely stabilize one axis while manipulating the other one/s
I also once came up with a customized PID controller that would accept properties, too - to hook up different PIDs dynamically.
But then someone in the FG community came up with a dedicated PID based autopilot that worked for most helicopters, so I considered that obsolete - until the maintainer of the corresponding helicopter decided the feature was unrealistic, and simply reverted the commits/contribution.
And like you say, hovering is a completely different beast (the generic AP I mentioned did support hovering).
Personally, I would probably "cheat" by using the groundspeed/position properties and hook those up to the PID
Again, it's possible to do this "as is", but the "modes" won't do what you expect them to do (as in heading or altitude hold), it's more akin to "fixing" one axis while manipulating the other ones.
I don't remember the details/name of the helicopter that had said generic rotary AP, but the commit logs should still contain all the nitty gritty details.
Like you, I would probably be inclined to implement something like that via Nasal code - but all that presumably happened at a time when property rules were not yet around, so there may be better ways today.