A computational model of cerebellum and midbrain to realize the human smooth constrained motion and bimanual cooperative motion

It is known that the intermediate part of cerebellum plays a key role in adjusting the motion of distal part of limbs to make the movement smooth. In this paper, a computational model for the intermediate part of cerebellum which ensures smooth contact motion with environment is developed. In this model, the cerebellum realizes a smooth contact motion by estimating the conditions of environment in the intermediate part of cerebellar hemisphere and tuning the desired motion via rubrospinal tract. The stability of contact motion is proven theoretically along the Lyapunov's direct method and its feasibility is demonstrated through some computer simulations. A redundant description of contact force using mirror symmetrically positioned two viscoelastic springs is introduced in the synthesis of contact motion controller. It makes the synthesis of consistent controller for both single arm contact motion and bimanual cooperative motion possible.