Many things have been done in this branch:
adjustement in the state-space computation so that now the mass matrix is also computed in the state space framework.
development of a Linear Quadrati Regulator (LQR) within a submodule: -- structure for LQR that allows the user to define the gain matrices -- implementation of Riccati equation solving -- inverse of a matrix via a LU decp -- schur factorization
The LQR submodule is using a state space representation to compute the appropriate gain. Once the feedback gain K is computed it can be used to apply the LQR control in user function (most probably in the user_joint_forces).