What representation is expected for gyro measurements? #9
I expected that any gyro measurement would instead be represented as Euler angles, that is, three orthogonal measurements of rotation rate. That's not algebraically equivalent to the axis-angle representation, as far as I can tell, except when any two of the components are 0.
What's the justification for this approximation? Is it expected that UAVs will usually only rotate around one body-frame axis at a time? Or perhaps you determined experimentally that the difference is small in practice?
The computation of
This is a standard representation for tactical grade strapdown systems. Also Euler angles are not orthogonal.