Task-Space Iterative Learning for Redundant Robots: Simultaneous Acquirements of Desired Motion and Force Trajectories under Constraints

This paper focuses on iterative learning control (ILC) for multi-joint robots with redundancy in degrees of freedom (DOFs) when task descriptions are given only in task-space. The specified task is of precise tracking of both endpoint position and force trajectories given on the plane by a five-DOF robot arm whose endpoint contacts with a plane during motion. Even in the case of twice continuously differentiable desired motion and force trajectories given only in task-space, it is possible to construct a learning update law in task-space by using the desired trajectories together with measured data correspond to them. The convergence of trajectory tracking under constraints by the proposed task space ILC is proved theoretically on the basis of nonlinear robot dynamics with DOF-redundancy.