search

PX4 / PX4-Autopilot

PX4 Autopilot Software
https://px4.io
BSD 3-Clause "New" or "Revised" License
8.26k stars 13.41k forks source link

Why not use airspeed sp for gain scaling if airspeed disabled? #11661

Open CarlOlsson opened 5 years ago

CarlOlsson commented 5 years ago

Currently we are scaling the actuators with the trim airspeed (a constant value) if airspeed is disabled in the fixed wing attitude controller https://github.com/PX4/Firmware/blob/b54ca67de9b5900c7ed11bef323c7b4e5a0339b6/src/modules/fw_att_control/FixedwingAttitudeControl.cpp#L480-L481

It seems sub optimal since the demanded airspeed can be significantly higher if the fixed wing position controller requests it due to ground speed undershoot in high winds. Is there a reason the scaling is not done using the demanded airspeed instead?

LorenzMeier commented 5 years ago

No, I think that's just historical. I think a test with demanded airspeed would be worthwhile.

Antiheavy commented 5 years ago

Sorry for my ignorance, but how can airspeed demand be reasonably estimated if airspeed is disabled in the controller?

I asked a related question in this other thread: https://github.com/PX4/Firmware/issues/11677

edit: I'm interested for a couple reasons: 1) we are starting to use both airspeed and non-airspeed modes, especially once this airspeed failsafe PR goes in: https://github.com/PX4/Firmware/pull/10733 2) we'd like any changes to fixed wing gain scaling to allow higher scaling below minimum airspeed (within reason) to allow more control authority during the immediate seconds after takeoff and before landing when airspeed could be below minimum.

CarlOlsson commented 5 years ago

how can airspeed demand be reasonably estimated if airspeed is disabled in the controller?

For some vehicles a reasonable estimate of airspeed can be acquired from the throttle setting scaled with air density and voltage

stale[bot] commented 5 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

stale[bot] commented 4 years ago

This issue has been automatically marked as stale because it has not had recent activity. Thank you for your contributions.

Antiheavy commented 4 years ago

Still might be interesting to look into this. FYI @Kjkinney

stale[bot] commented 4 years ago

This issue has been automatically marked as stale because it has not had recent activity. Thank you for your contributions.