* added support for resetting wind states to external observation
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* moved wind related functions into separate file
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* added VEHICLE_CMD_EXTERNAL_WIND_ESTIMATE
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* correctly compute variances
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* ekf2: implement wind reset
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* only allow VEHICLE_CMD_EXTERNAL_WIND_ESTIMATE during wind dead-reckoning
and increase horizontal velocity variance to allow velocity states to move
towards solution that is aligned with the newly set wind
Signed-off-by: RomanBapst <bapstroman@gmail.com>
* only reset wind on ground
* still use wind reset using airspeed when it wasn't initialized
* exclude func for rover, change reset interface
* handle wind reset in drag-fusion
* replace state reset with variance reset in sideslip/drag fusion
* remove resetWind function
---------
Signed-off-by: RomanBapst <bapstroman@gmail.com>
Co-authored-by: Marco Hauswirth <marco.hauswirth@auterion.com>
* ekf2_derivation: use single source of state definition
The state is defined as an ordered dictionary of group elements and
everything else is generated using that state definition
* ekf2: generated state sample add const reference getter
---------
Co-authored-by: bresch <[brescianimathieu@gmail.com](mailto:brescianimathieu@gmail.com)>
Co-authored-by: Daniel Agar <daniel@agar.ca>
* ekf2-test: remove outdated codegen comparison
The definition of states changed so the comparison with the old
derivation cannot work anymore.
---------
Co-authored-by: bresch <brescianimathieu@gmail.com>
- from the C standard, "All identifiers that begin with an underscore and either an uppercase letter or another underscore are always reserved for any use"
ekf2: merge mag 3d innov var, Hx and Kx computation to reduce flash
Slightly less code produced, almost no performance change
ekf2_mag3D: do not pre-compute Kalman gains
The vector of Kalaman gains is not too expensive to compute using
matrix-vector multiplication. Pre-generating it using CSE takes a lot of
flash space for little benefit.
bresch