When an airspeed sensor is not used, during a takeoff, the pitch angle
is asymptotically driven to 0 as the takeoff altitude is approached.
Some airplanes will then stop climbing and fail to reach altitude.
To prevent an indefinite wait for the takeoff altitude to be reached, a
dedicated level-off timeout has been introduced.
- TAKEOFF and AUTO flight modes now should have identical takeoff
- Prevent behaviour switching past climb altitude in TAKEOFF mode.
- Refactor set_pitch_min/max methods.
Max was already there, now renamed.
Min is newly introduced.
behaviour.
- Remove enforcement of min takeoff throttle logic from servos.cpp.
It is now handled only by takeoff.cpp.
- Take TKOFF_LVL_ALT into consideration in AUTO as well.
- Fixed pitch setpoint when TKOFF_ROTATE_SPD>0.
- Roll navigation in mode TAKEOFF during climb should now work again.
- Now the TAKEOFF loiter waypoint is set by the bearing of the
aircraft while on TKOFF_LVL_ALT, as in the last stable release, instead
of TKOFF_ALT.
- Using TRIM_THROTTLE in takeoffs, when TKOFF_THR_MIN==0
this is principally for tailsitters where rangefinders would be
orientation with RNGFND1_ORIENT=12 (PITCH_180), but also allows for
custom orientations which will be useful if the rangefinder is tilted
forward
This is a rework so that servos.cpp is responsible for setting the
throttle limits under more circumstances and always notifies TECS when
it does so.
Additionally, the TAKEOFF mode has been improved with a new parameters
TKOFF_MODE and TKOFF_THR_MIN that extend the throttle behaviour.
This variable updated unpredictably, and it was easy to introduce bugs.
It was not used in many places and is clearer to calculate the error
directly when needed.
* Set position target used to just be used in scripting, now it's used
by DDS in external control
Signed-off-by: Ryan Friedman <ryanfriedman5410+github@gmail.com>
this fixes a bug where if the terrain database cache does not have the
tile for the location of a rally point then RTL to the rally point
with TERRAIN_FOLLOW=1 will not track terrain
The underlying issue is that Location::loc.change_alt_frame() will
return false if the location is not in the terrain memory cache. We
can't just extrapolate as the rally point could be in a totally
different terrain area to the current location. So instead we set it
as terrain_following_pending and fix it as soon as the terrain cache
is filled.
fixes https://github.com/ArduPilot/ardupilot/issues/25157
this fixes a bug where a change of home altitude would cause a sudden
height demand change. This copes with 3 situations:
- flying with AMSL alt demand. Changing home altitude makes for no change
- flying with AGL alt demand. Changing home altitude requires update of next_WP_loc
- flying with home relative alt demand. Changing home altitude changes demand at end of current navigation leg
George Zogopoulos
Michelle Rossouw
Iampete1
Peter Barker
timtuxworth
Andrew Tridgell
Henry Wurzburg
Bob Long
Tom Pittenger
Ryan Friedman
Mykhailo Kuznietsov