A scaling algorithm for the generation of jerk-limited trajectories in the operational space

Kinematic singularities represent a relevant problem for trajectories that are defined in the operational space. In case of industrial applications characterized by non-repetitive tasks, feasibility cannot be checked in advance, so that appropriate methods have been developed for the online management of otherwise critical situations. In this paper, a scaling scheme proposed in the past for the automatic handling of possibly unfeasible trajectories is revised in order to generate jerk-limited reference signals: close to critical points, trajectories are appropriately slowed down such to guarantee an accurate tracking of the assigned path in the operational space. The actual performances of the proposed system have been experimentally verified on a commercial manipulator by means of extensive tests.