Nav rotorcraft rework #2964
Nav rotorcraft rework #2964
Conversation
|
Thx, nice development. @gautierhattenberger @Fabien-B Small remark, It would be great if the testdrone color is not the same green as the expected track, can you make it Pink, Yellow or Reddish, or Cyan please. |
There was a problem hiding this comment.
Nice work Gautier! I think it was a great idea to bring a bit more structure to the guidance. I had a few questions looking through the code, but from what I could see it all made sense. It's hard to keep track of all the changes, as some things were just moved. I'll try to do some simulation tests soon!
sw/airborne/firmwares/rotorcraft/stabilization/stabilization_indi.c
Outdated
Show resolved
Hide resolved
sw/airborne/firmwares/rotorcraft/guidance/guidance_indi_hybrid.c
Outdated
Show resolved
Hide resolved
it should be already done before in guidance, depending if the flag GUIDANCE_HEADING_IS_FREE is set or not
|
@EwoudSmeur after testing again in simulation, I realized that the |
|
|
||
| <dl_setting var="cruise_throttle" MIN="0" STEP="1" MAX="9600" shortname="cruis_throttle"/> |
There was a problem hiding this comment.
I have removed the shortname instead of correcting the spelling since it was the same than name (so it makes no difference in the end)
|
After maybe @EwoudSmeur has checked it we can merge it. |
| @@ -9,7 +9,7 @@ | |||
| <settings> | |||
| <dl_settings> | |||
| <dl_settings NAME="guidance_indi_hybrid"> | |||
| <dl_setting var="gih_params.liftd_p50" min="0.1" step="0.1" max="10.0" shortname="liftd_p50" param="GUIDANCE_INDI_LIFTD_P50" persistent="true"/> | |||
| <dl_setting var="gih_params.liftd_p50" min="0.1" step="0.1" max="10.0" shortname="liftd_p50" param="GUIDANCE_INDI_LIFTD_P50" persistent="true" module="guidance/guidance_indi_hybrid"/> | |||
There was a problem hiding this comment.
Why is there a module added here?
There was a problem hiding this comment.
The module attribute is a legacy prior to the "modules" we have now. It can be replaced by header and it was used to specified the header file for settings.
We will clean that someday...
| &(stateGetPositionEnu_i()->x), | ||
| &(stateGetPositionEnu_i()->y), | ||
| &(stateGetPositionEnu_i()->z), | ||
| &pos->x, |
There was a problem hiding this comment.
If our navigation is in floats does it still make sense to send it down in integers? (Maybe something for later to update)
There was a problem hiding this comment.
yes, indeed, but it is more related to redefining most messages, so it is an other job
This is a complete rework of the navigation for rotorcraft with the key features: - all nav functions and interface in float - nav function are registered (decoupling between nav API and implementation of standard pattern) - submodes to specify if setpoint is a position (legacy), speed or accel (better integration of algorithms like GVF) - guidance (H and V) is reorganized with default function to implement (run_pos, run_speed, run_accel) - guidance control (the old default PID) is separated from the guidance logic --------- Co-authored-by: Freek van Tienen <freek.v.tienen@gmail.com>
This is a complete rework of the navigation for rotorcraft with the key features:
Still some questions ahead:
@EwoudSmeur @fvantienen What is your opinion about this approach ? Do you see a proper way to integrate the indi/hybrid in a with this ?
I would like to split the stabilization from guidance, but maybe it is a bit to much for a single pull request ?