Robot Control based on Skill Primitives

M. Kolbus, T. Reisinger, and J. Maaß (Germany)


Control, Hybrid Control, Motion planning, Parallel Kine matic Machines


Parallel kinematic robots need model based control to provide superior dynamic capabilities. To execute assem bly tasks defined by skill primitives, a self-configuring hybrid control architecture featuring jerk-limited end effector trajectories is required. In this paper control struc tures taking these special demands into account are pre sented. Focus is put on the theory of a modified computed torque controller and an algorithm for time-optimal mo tion planning respecting the task frame formalism. The proposed concept is supported by promising simulation results and it is supposed to lead to superior overall per formance in the execution of assembly tasks by parallel kinematic robots.

Important Links:

Go Back