Fig 1.
GMM-KVS method flowchart.
Fig 2.
Schematic diagram of the elbow method.
Fig 3.
Schematic diagram of the elbow method (when the effect is poor).
Fig 4.
Theoretical demonstration trajectories: (A) system 1; (B) system 2; (C) system 3.
Fig 5.
Line chart plotted using the elbow method: (A) system 1; (B) system 2; (C) system 3.
Fig 6.
Line chart plotted using the k-value selection algorithm: (A) system 1; (B) system 2; (C) system 3.
Fig 7.
Trajectory comparison for system 1: (A) starting point within collection range; (B) starting point outside collection range.
Fig 8.
Trajectory comparison for system 2: (A) starting point within collection range; (B) starting point outside collection range.
Fig 9.
Trajectory comparison for system 3: (A) starting point within collection range; (B) starting point outside collection range.
Fig 10.
Comparison of error values for trajectories generated by DMP, GMM, and GMM-KVS: (A) mean absolute error; (B) root mean square error.
Table 1.
Evaluation values of DMP, GMM, and GMM-KVS (simulation experiment).
Fig 11.
Demonstration process of the robotic arm transporting an object: (A) starting position; (B) via position 1; (C) via position 2; (D) target position.
Fig 12.
Collected demonstration trajectories.
Fig 13.
Line charts of the elbow method and the k-value selection algorithm: (A) elbow method; (B) k-value selection algorithm.
Fig 14.
Trajectory comparison chart: (A) starting point 1; (B) starting point 2; (C) starting point 3.
Table 2.
Evaluation values of DMP, GMM, and GMM-KVS (robotic arm experiment).
Fig 15.
The process of the robotic arm autonomously transporting an object: (A) starting position; (B) via position 1; (C) via position 2; (D) target position.