Document semantics difference between Pinocchio quantities and iDynTree KinDynComputations quantities

traversaro commented 4 years ago

DOFs and link serializations
Velocity, force and acceleration representations for the base quantities
How we need to transform the vector and matrix input and output quantities appropriately given these differences

traversaro commented 4 years ago

cc @GiulioRomualdi

traversaro commented 4 years ago

Related issues:

GiulioRomualdi commented 4 years ago

Before I forgot I want to drop here this table of conversion:

`iDynTree`	`Pinocchio`
MIXED_REPRESENTATION	LOCAL_WORLD_ALIGNED
INERTIAL_FIXED_REPRESENTATION	WORLD
BODY_FIXED_REPRESENTATION	LOCAL

DanielePucci commented 4 years ago

CC @robotology/iit-dynamic-interaction-control

traversaro commented 2 years ago

Before I forgot I want to drop here this table of conversion:

Updated version to be self-contained. Assumptions:

we have two frame a and e, and the transform between this two frames is represented by ᵃRₑ ∈ SO(3) ᵃoₑ ∈ ℝ³
(.)ᵛ represents the mapping between the skew-symmetrix matrix and the related vector in ℝ³, i.e. such that if u,v ∈ ℝ³, u = (U)ᵛ s.t. U v = u × v
the dot represent the usual mathematical analysis dot operator (time derivative) and not the Featherstone one then the different velocities are:

`iDynTree`	`Pinocchio`	Math
MIXED_REPRESENTATION	LOCAL_WORLD_ALIGNED	(ᵃȯₑ , ( ᵃṘₑ ᵃRₑᵀ )ᵛ ) .
INERTIAL_FIXED_REPRESENTATION	WORLD	(ᵃȯₑ - (ᵃṘₑ ᵃRₑᵀ) ᵃoₑ , ( ᵃṘₑ ᵃRₑᵀ )ᵛ ) .
BODY_FIXED_REPRESENTATION	LOCAL	(ᵃRₑᵀ ᵃȯₑ , ( ᵃRₑᵀ ᵃṘₑ )ᵛ ) .

Sources and derivation for the equations:

Related Pinocchio issues:

robotology / idyntree