Optimal Trajectory Planning of Industrial Robots using Geodesic

Abstract

This paper intends to propose an optimal trajectory planning technique using geodesic to achieve smooth and accurate trajectory for industrial robots. Geodesic is a distance minimizing curve between any two points on a Riemannian manifold. A Riemannian metric has been assigned to the workspace by combining its position and orientation space together in order to attain geodesic conditions for desired motion of the end-effector. Previously, trajectory has been planned by considering position and orientation separately. However, practically we cannot plan separately because the manipulator joints are interlinked. Here, trajectory is planned by combining position and orientation together. Cartesian trajectories are shown by joint trajectories in which joint variables are treated as local coordinates of position space and orientation space. Then, the obtained geodesic equations for the workspace are evaluated for initial conditions of trajectory and results are plotted. The effectiveness of the geodesic method validated through numerical computations considering a Kawasaki RS06L robot model. The simulation results confirm the accuracy, smoothness and the optimality of the end-effector motion.