Learning control scheme for a class of robot systems with elasticity

A new type of iterative learning control scheme is proposed for robot manipulators composed of rigid links and driven by actuators through transmissions with elasticity. This method is based on the learning scheme called "betterment process" for operation of a mechanical robot in a sense that it betters the next operation of a robot by using previous operation's data. In the first stage, a betterment process for the rigid link subsystem is employed in order to realize a desired motion pattern by regarding the interconnected terms with the other drive subsystem of actuators as hypothetical inputs. In the second stage, another betterment process is employed to make the hypothetical inputs in terms of the control torques for actuators. The convergence of this "2-stage betterment process" to the desired motion trajectory is assured under some reasonable conditions. To show the effectiveness of the proposed learning scheme, some partial results on practical applications of this method for an actual robot manipulator, together with numerical results by computer simulation, are given.