Time-Optimal Trajectory Planning for a Pu’er Tea Feeding Manipulator Based on the SA-PSO Algorithm

Robotic feeding in tea production requires time-efficient joint-space trajectories that satisfy kinematic constraints while maintaining continuity at prescribed task configurations. This study proposes a simulated annealing-integrated particle swarm optimization (SA-PSO) algorithm for time-optimal trajectory planning of an ABB IRB 120 manipulator used in Pu’er tea feeding. The proposed approach couples the optimization of intermediate joint configurations with uniform time scaling of a 3–5–3 piecewise-polynomial trajectory. A shared temperature schedule coordinates the adaptation of PSO learning factors and the Metropolis acceptance criterion, while separate archives retain the best feasible solutions found during the search. The method was evaluated in MATLAB and compared with four other optimization algorithms over 20 independent runs per algorithm. SA-PSO achieved a mean planned motion time of 9.125913 s. Compared with conventional particle swarm optimization (PSO), it reduced the mean planned motion time by 0.292 ms and the sample standard deviation of planned motion time by 98.56%. Robustness was further assessed by perturbing the internal waypoints shared by adjacent trajectory segments. Across these tests, all 70 generated trajectories remained feasible under Gaussian noise levels of up to 0.20 rad. These results indicate that the proposed method produces consistent planned motion times across independent runs and maintains trajectory feasibility under the tested waypoint perturbations.