Adaptive Time-Optimal Trajectory Planning Subject to Kinematic and Dynamic Constraints for Articulated Robots

• Published in: IEEE robotics and automation letters Vol. 10; no. 4; pp. 4085 - 4092
• Main Authors: Gan, Yahui, Xu, Jiewei, Cai, Wei, Huang, Bin
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.04.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Constrained motion planning; Constraints; Controllability; Convexity; Dynamics; Friction; Heuristic algorithms; Kinematics; motion control; Parameterization; Planning; Reachability analysis; Robots; Torque; Training; Trajectory optimization; Trajectory planning
• ISSN: 2377-3766; 2377-3766
• DOI: 10.1109/LRA.2025.3549647

This letter proposes an adaptive time-optimal path parameterization algorithm (A-TOPP) subject to kinematic and dynamic constraints for articulated robots. The algorithm efficiently incorporates the effects of Coulomb and viscous friction, while flexibly and comprehensively addressing various kinematic and dynamic constraints in different task scenarios. It emphasizes the simultaneous management of both kinematic and dynamic constraints, ensuring versatility and high efficiency across various operational contexts. Firstly, reachable and controllable curves are generated bidirectionally based on reachability analysis. When the maximum velocity curve is exceeded, the numerical search is conducted downward to find the maximum velocity point. Subsequently, local bidirectional adaptive computations are performed to self-correct controllable and reachable curves, thereby obtaining the optimal velocity curve. Finally, the proposed method is validated using a 6-DOF robot along predetermined geometric paths. The results indicate that the A-TOPP algorithm effectively addresses the limitation of the time-optimal path parameterization approach based on reachability analysis (TOPP-RA) which only handles the single torque constraint in ideal cases. Its planning precision is nearly identical to that of the convex optimization algorithm (TOPP-CO) on different paths and remains highly consistent even under various constraints. Notably, compared to the TOPP-CO algorithm, the computational efficiency of the A-TOPP algorithm has soared dozens of times, firmly attesting to its remarkable efficacy.