Near Time-Optimal Flexible-Joint Trajectory Planning Algorithm for Robotic Manipulators

Abstract

In this paper, a new hybrid time-optimal flexible-joint trajectory planning algorithm is introduced. Time-optimal rigid-body trajectories contain sudden changes of acceleration, which can result in violation of constraints when applied to flexible robotic manipulators. The abrupt acceleration changes are replaced by a smooth time-optimal switching strategy, which is based on the solution of a two-mass oscillator two-point boundary-value problem. It is designed such that the two-mass oscillator's flexible modes are not excited. In between the transitions, a multi-axis rigid-body model is used. This combination of flexible and rigid-body models allows to design a computationally very efficient trajectory planning algorithm, which considers multi-axis dynamics as well as some important inherent flexibilities.