Pinocchio: Floating base robot q and v

[xinsongyan]

Hi,

I am using pinocchio to calculate a floating base robot kinematics and dynamics, but I am not sure about the convention of _q_ and _v_?

I used RBDL before, the _q_ and _v_ in RBDL are defined as:

q = [global_base_position, global_base_quaternion_x, global_base_quaternion_y, global_base_quaternion_z, joint_positions, global_base_quaternion_w]
v = [global_base_velocity_linear, local_base_velocity_angular, joint_velocities]

I tried a little bit pinocchio, it seems that it takes:

q = [global_base_position, global_base_quaternion, joint_positions]
v = [local_base_velocity_linear, local_base_velocity_angular, joint_velocities]

I didn't find any documentation about this. Anyone can confirm this?
Thanks a lot!

[jcarpent]

You're correct. The base translation part is expressed in the parent frame (here the world coordinate system) while its velocity is expressed in the body coordinate system.