Write control surface configs to model results

Jaeyoung-Lim commented 3 years ago

Problem Description This PR writes the control surface configs to the model results so that it can be parsed from the gazebo simulation

Questions

I am not sure how we can general control surfaces: do we need to know the axis of the deflection? or do we want to have all three axis included in the coefficients?

Jaeyoung-Lim commented 3 years ago

Merging, since I think this is obvious.

manumerous commented 3 years ago

Is there a particular reason we need the separate wing category? And why do we put the tail control surfaces like elevator and rudder under the wing category?

I was planning to separate the wing and tail components completely at some point for all models.

Jaeyoung-Lim commented 3 years ago

My intention was to have the exact same structure as rotors. Control surfaces before did not have any category, therefore needed to be handled differently. I added so that it is more similar. I would be even happier if we can even mix motors and control surface definitions in the model config - completely removing categories and simply defining actuator type. But I think this will become quite hard with tiltwings so leaving this for future work.
Since we are using a lumped-sum model there is not much value IMO separating tail surfaces and wing control surfaces (they do the same thing). Therefore I was trying to put everything under the wing category and assume that a vehicle is a single lifting surface
Probably we do not want to remove the categories if we want to support a multi element model that can also take into consideration leaver arm moment effects.

I was planning to separate the wing and tail components completely at some point for all models.

Great, if this was the plan there was no reason to wait :wink: However, I think we need to discuss at some point how we want to represent control surfaces.

manumerous commented 3 years ago

Okay that makes sense I guess. The advantage of separating them would be in my opinion that you could use the standard wing model also for delta planes and that it does not come with a tail in every case.

Yes I think it might be better to discuss that in a short meeting. :)

Jaeyoung-Lim commented 3 years ago

@manumerous Whenever you are available :grin:

ethz-asl / data-driven-dynamics

Write control surface configs to model results #133