Add new flight mode: Airspeed (only for fixed-wing flight) - continued

[Jaeyoung-Lim]

Describe problem solved by this pull request
#17924 continued with the new structure of #18340 included

Test data / coverage

• Tested in SITL Gazebo

make px4_sitl gazebo_plane

Additional context

• This PR is a continuation of [WIP] Add new flight mode: Airspeed (only for fixed-wing flight) #17924
• This PR will be rebased when Encapsulate position setpoint and loiter setpoint handling for FW position controller #18340 is merged

[Jaeyoung-Lim]

@dagar Is there something wrong with caching on github actions? This PR builds and runs on my local machine, but somehow fails on github actions

[dagar]

@dagar Is there something wrong with caching on github actions? This PR builds and runs on my local machine, but somehow fails on github actions

I hope not, but I'll take a look...

[dagar]

I'm seeing the same build failures locally.

[Jaeyoung-Lim]

@dagar It took a while that that header was being generated and I was editing a generated file. Sorry about the confusion

[Jaeyoung-Lim]

Rebased due to conflict

[kaklik]

Hi! This new flight mode is actually what I looking for in the case of an unpowered flight. Would be nice, if it could be merged soon.
Thanks.

[Jaeyoung-Lim]

@kaklik We need a flight test for this to be merged. Would you be interested on testing this?

[dagar]

Steal the legacy rattitude slot?

[kaklik]

@kaklik We need a flight test for this to be merged. Would you be interested on testing this?

Yes, but to clarify the situation. I want to use it with the autogyro. Therefore I am unable to test it with "fixed-wing" exactly.

[roman-dvorak]

Hi @Jaeyoung-Lim, I think we are able to do such a flight test with our TF-G2 autogyro. (TF-G2 is currently controlled by FW controllers)

We would need a few things to do that:

• Resolve new conflicts against to the current master version
• Tell me how to run this mode .. Is it already supported by QGC? Or do we have to send our own mavlink message that switches flight mode?
• Are there any specific requirements for such test?

[dagar]

Should we only allow entering this mode if airspeed is currently valid?

[sfuhrer]

I think we should yes, it's of no use in case of invalid airspeed.
We can also think about using it as failsafe mode from Altitude mode in case of invalid altitude estimation.

[roman-dvorak]

Hi, I assume the line 2272 should be bool in_air_use_tecs = ...?

PX4-Autopilot/src/modules/fw_pos_control_l1/FixedwingPositionControl.cpp

Lines 2270 to 2282 in 0a5c249

	/* tell TECS to update its state, but let it know when it cannot actually control the plane */
	bool in_air_alt_control = (!_landed &&
	(_control_mode.flag_control_auto_enabled ||
	_control_mode.flag_control_velocity_enabled ||
	_control_mode.flag_control_altitude_enabled ||
	_control_mode.flag_control_climb_rate_enabled ||
	_control_mode.flag_control_airspeed_enabled));
	/* update TECS vehicle state estimates */
	_tecs.update_vehicle_state_estimates(_airspeed, _body_acceleration(0), (_local_pos.timestamp > 0), in_air_use_tecs,
	_current_altitude, _local_pos.vz);

[roman-dvorak]

Hi @Jaeyoung-Lim we plan to do some flight tests tomorrow. And we could test this airspeed mode.

But I can't enter intu airspeed mod in simulator (in FlightGear). Specifically, it was this commit

I tried to takeoff in the simulator and do the command in flight:

pxh> 
pxh> commander mode airspeed
pxh> main_ret: -1
INFO  [tone_alarm] notify negative

How I can enter into flight mode?

[roman-dvorak]

Hi, we made two successful attempts with airspeed mode. Booth tests were done with TF-G2 autogyro.

First try

There was a logging issue on the first try. So we don't have a log.

It was very interesting to control such a setting and it gave the dron completely different properties. For certain flights, such flight mode certainly make sense. Switching to this mode was done from the command line with the command commander mode airspeed.
The flight in this mode lasted over 10 minutes and several horizontal flights were performed, several climbing to higher altitudes and subsequent gliding without motor was done. No unexpected behaviour was observed during the flight. Drone is quite "action" and adroit in it.

Second try

During the test, we wanted to do more partial tests (because it was a long time after which we have a log) for example, automatic autogyro takeoff (#18582) and other test verifying the new properties of the autogyro. Therefore, the flight in airspeed mode was short. It was twice a climb and a motorless descend.

No problem with airspeed was observed this time either. And the controls were pretty good.

log: https://review.px4.io/plot_app?log=be4b0ce6-b51c-4826-907d-f980df96fbc1
(This is the first flight in a long time, from which we have a proper log :) )

I noticed strange things in the log, that the use of RAM at the time of switching to airspeed mode has increased a bit.

Next usage

In the near future, we would like to use this flight mode to launch our TF-G2 autogyro from a stratospheric balloon. Without a motor. This mode will be suitable for such an experiment.