[control] rework the stabilization for rotorcraft #3248
Conversation
gautierhattenberger
requested review from
dewagter,
fvantienen,
EwoudSmeur and
Fabien-B
|
It seems that the RC inputs are not correctly handled anymore. I'm looking at this. |
gautierhattenberger
force-pushed
the
stabilization_rework-integration
branch
from
4c8e983
to
c2069db
Compare
|
Updating with a better support of RC inputs, where each part is properly separated (AP, stab, guidance_h/v) |
gautierhattenberger
force-pushed
the
stabilization_rework-integration
branch
from
c2069db
to
1c41caf
Compare
Introducing several changes to the control stask for the stabilization: - new structures: stabilization, stab (att and/or rates) and thrust (vector) - more functional approach: stabilization receives the stab and thrust setpoints from guidance and fill command vector at the end - modes and sub-modes are redefined (split stab and guidance modes), some AP modes are removed (FLIP and MODULE) - RC read functions are removed and replaced by ABI bindings, one for each part (AP, stab, guidance_h, guidance_v) with dedicated functions - transition ratio (float in [0-1]) replace transition_percentage (int) - the AP_MODULE mode is replaced by generated autopilot - the ABI hack for guidance indi is not needed anymore - AP generator supports a new 'store' attribute to use return values of functions - cleaning of the stabilization API (for setter functions)
gautierhattenberger
force-pushed
the
stabilization_rework-integration
branch
from
1c41caf
to
0d42cd1
Compare
|
@EwoudSmeur @fvantienen @dewagter, did you get a chance to test this branch ? I would like to integrate this soon, by the end of the month if possible. Thanks |
There was a problem hiding this comment.
Hi Gautier, thanks for your hard work! I simulated some of our drones and that went smoothly! I also looked through the code, I don't really have comments because it is hard to see the structural changes because of all the refactoring, but I agree with your idea. Thank you!
|
@gautierhattenberger: this nice but big PR #3248 together with Freek's #3254 for the black cube chibi and Tomaso's current #3261 and next updates (#3237) will contain quite important changes. The current version in Would it be an option to make a new release first and merge these big changes in a fresh dev branch? |
|
The last stable version dates from December, with around 30 commits since that. Next release is usually around June, but if you feel that the changes are worth making a release, it is possible. Honestly, we have not flown a lot during the last months... Note that the current tag is "v7.0_unstable" :) |
|
@dewagter so, do you prefer a 7.0 stable version before merging this ? |
Introducing several changes to the control stask for the stabilization:
Most of the changes should have been tested in simulation, but not much in real flight. Even if it is a quite big changes, I think it is an overall simplification of the control architecture.