Extend the IK to implement the hierarchical quadratic program (HQP) technique

[GiulioRomualdi]

It would be nice to design a new concrete implementation for the IK to consider hierarchical quadratic program as in https://hal.archives-ouvertes.fr/hal-00751924/document.

As far as I know there has been no open source IK-HQP solver available. https://github.com/tasts-robots/pink and https://github.com/stephane-caron/pymanoid tries the emulate the prioritization by choosing a specific set of weights in the cost function (that's similar to what we have in blf)

Thanks to this it will be possible to clearly exploit the task prioritization in the framework, and mimic the null space jacobian task priority method

Additional info: https://scaron.info/robot-locomotion/inverse-kinematics.html

[GiulioRomualdi]

This work is also interesting: https://hal.archives-ouvertes.fr/hal-00486755v2/document