Replication data for: Planning of Curvature-Optimal Smooth Paths for Industrial Robots Using Neural Networkshttps://doi.org/10.18419/darus-2126Kaiser, BenjaminDaRUS2021-09-142021-09-14T15:02:59ZThis dataset contains curvature-optimal geometry parameters for linear-linear transitions for polynomial smoothing with polynomial splines 4th order.
The data was generated by offline solving the optimization problem of curvature-optimal geometry parameters for corner smoothing, which is described in the related publication. Dataset_linlin.tab contains optimal geometry parameters for linear-linear transitions. Dataset_lincirc.tab contains the optimal geometry parameters for linear-circular transitions.
These datasets were used to train the neural networks as described in the corresponding publication.
cartesian_trajectory_non_smooth.tab contains the non-smoothed example trajectory from the corresponding publication. The orientations are RPY (xyz) Euler angles. The velocity is planned to keep the jerk limits.
cartesian_trajectory_smooth.tab contains the smoothed example trajectory from the corresponding publication. The orientations of the trajectory are RPY (xyz) Euler angles. The trajectory was planned using the neural network to determine the optimal geometry parameters. The velocity is planned to keep the jerk limits.
cartesian_trajectory_fs.tab contains the non-smoothed example trajectory from the corresponding publication. The orientations are RPY (xyz) Euler angles. The velocity profile of the trajectory ignores the jerk boundaries.
trajectory_non_smooth, trajectory_smooth and trajectory_fs are the corresponding trajectories in the jointspace. These were planned for the Kuka KR500 R3330 with the end effector from the corresponding publication.EngineeringCorner smoothingcurvaturepolynom splineB. Kaiser and A. Verl, "Planning of Curvature-Optimal Smooth Paths for Industrial Robots Using Neural Networks," 2021 Fourth International Conference on Artificial Intelligence for Industries (AI4I), 20212021-06-15Kaiser, Benjamin2021-09-07CC BYCC BY Waiver