Robot Force/Position Control Combined with ILC for Repetitive High Speed Applications

In this paper an approach for robot force/position control combined with an iterative learning control is proposed. Following high speed force trajectories in different repetitive robotic applications is a challenging field in robotics. Such applications require a desired contact force while following a position/orientation trajectory in the non-force controlled directions. For this a parallel force/position control is suitable, but when it comes to high speed tasks with varying contact stiffness along the trajectory such a method reaches its dynamical limit. The problem can be solved by using the parallel force/position control to learn the trajectory for a slowed down task and correct this trajectory step by step towards the original task speed. When the original task speed is reached an iterative learning control law combined with the force/position control is used to further reduce the force error.