Enhanced calibration algorithm robust to many redundancies and singularities of kinematic model tested on 3[RRPaR] Delta robot

Here we present the data associated with our paper "Enhanced calibration algorithm robust to many redundancies and singularities of kinematic model tested on 3[RRPaR] Delta robot": "desired_poses_mm" contains 160 desired points in the robot workspace. Each row corresponds to the Cartesian coordinates [x, y, z] of one point, expressed in millimeters in the laser tracker coordinate frame {L}. These points form the cuboid with its spatial diagonals (shown in Fig. 10). "measured_poses_nominal_model_mm" contains the positions of the corner reflector measured by the laser tracker when the robot was commanded to the desired points in "desired_poses_mm" using the nominal (uncalibrated) model. Each row corresponds to the measured Cartesian coordinates [x, y, z], expressed in millimeters in the {L} frame. "measured_poses_simplified_model_mm" contains the positions of the corner reflector measured by the laser tracker when the robot was commanded to the same desired points using the calibrated simplified model (Section 5.1). Each row corresponds to [x, y, z] coordinates in millimeters in the {L} frame. "measured_poses_generalized_model_mm" contains the positions of the corner reflector measured by the laser tracker when the robot was commanded to the same desired points using the calibrated generalized model (Section 5.2). Each row corresponds to [x, y, z] coordinates in millimeters in the {L} frame. For each of the three models (nominal, calibrated simplified, and calibrated generalized), the pointwise deviations between the desired and measured positions were computed across all 160 target points. These differences are plotted in Fig. 14.