An iterative learning control technique for a dual arm robotic system

A multidimensional iterative model for the control analysis of two cooperating robots is proposed. The manipulation of a rigid object with elastic restraint between the object and the end effectors using the proposed model is investigated. The scheme is based on an iterative structure which utilizes available data from the previous operations. These data consist of positions and velocities of the actual trajectories and applied forces to the object at the points of contact. It is shown that the convergence of the actual trajectory for the manipulation of the rigid object to the desired one is guaranteed upon proper selection of two weighted control matrices.<
>