Smooth and Accurate Trajectory Planning for Industrial Robots

Abstract

This paper proposes a smooth and accurate trajectory planning for industrial robots using geodesics. The workspace of a robot is split into positional and orientational parts. A Riemannian metric is given on each space such that the desired motion is a geodesic for the given metric. By regarding joint variables as local coordinates of the position space and the orientation space, Cartesian trajectories are represented by joint trajectories. A smooth and accurate motion of the robot end-effector and smooth joint trajectories corresponding to the motion can be obtained by calculating geodesics on the position space and the orientation space. To demonstrate the effectiveness of the proposed method, simulation experiments are conducted using the PUMA 560 robot.