mTECS branch wind estimation

[LorenzMeier]

This adds logging of the wind estimate. Compiles, not tested in HIL / flight yet.

[LorenzMeier]

@thomasgubler @julianoes Review would be appreciated. It would be good to have a larger number of logs to see if the wind speed estimation is reliable and can be used. Not high on the priority to fix it, but we need a lot of data, so this should go in soon-ish.

[thomasgubler]

Why are you defining wind speed only as a 2d velocity instead of 3d?

[LorenzMeier]

We can't estimate it in 3D with a pitot facing only into the plane and movements that are almost exclusively in 2D.

[thomasgubler]

Well, not without a good model...
Anyway, but why not define the communication interface (uorb) in a more generic way?

[LorenzMeier]

Because we should extend interfaces as they become relevant and not create empty fields that we have no clue yet how to use. There is no reliable measurement in Z direction and you can go as far as to argue that its not observable without forcing the system into a very special motion pattern and/or add 3D airspeed sensing. Both is not relevant for at least months if not longer and adding fields is possible ad-hoc (the log format is self-describing).

[thomasgubler]

Tested in HIL, no analysis of correctness of estimate done, but no operational problems found, merging