I do not record the raw data from the sensor so this is the best I can provide.

I have three rotation angles theta, phi and psi in radians.

Rotation is relative to an x axis pointing in the heading direction of the robot.  y is to the robot's left.  z is up in the robot frame.

The theta is the first rotation about the z-axis.

phi is next about the y-axis.

Then comes psi about the x-axis.
 
The timestamps are in seconds and have been adjusted to agree with the odometer timestamps.

The data has also been slightly lowpass filtered to remove noise.

The variance in the theta angle seems to be about:

P=1E-6*(time interval) (rads)^2
 
 

Back to main page  John Folkesson johnf@nada.kth.se