Fix documentation of KinDynComputations::inverseKinematics

[traversaro]

KinDynComputations::inverseKinematics ( https://robotology.github.io/idyntree/classiDynTree_1_1KinDynComputations.html#a60870f0f43c035fc63527ad42394f963 ) documentation is confusing w.r.t. the frame used to represent each element of the input linkExtForces . The documentation just reports:

The semantics of baseAcc, the base part of baseForceAndJointTorques and of the elements of linkExtWrenches depend of the chosen FrameVelocityRepresentation .

that does not mean anything. Even worse, if one clicks on LinkNetExternalWrenches, the link brings to https://robotology.github.io/idyntree/classiDynTree_1_1LinkWrenches.html, where it is written:

In both cases the Wrench corresponding to the link with LinkIndex i is always expressed with the orientation of the link frame and with respect to the link frame origin.

That may be true for iDynTree::LinkWrenches, but for sure it is not true for iDynTree::LinkNetExternalWrenches when use in the context of KinDynComputations::inverseKinematics related methods. We should fix the documentation to avoid these confusion. cc @AntonellaTatarelli @claudia-lat