A new control method applied to robots with joint elasticity

This paper presents an application example of a novel disturbance-rejection control method. The example considers a flexible-joint robot where actuator dynamics is not ignored. After a review and a discussion of the state of art in controling flexible-joint robots, the novel method is derived for the general case and then applied to the considered example. The key feature of this method is to transform nonlinear effects (disturbances) acting on the system to where they can be estimated and statically compensated. For these purposes, only a rough model knowledge and only the position measurement are required. Simulation results show the superiority of the proposed method.<
>