To improve the safety and effectiveness of autonomous towing aircraft aboard the carrier deck, this study proposes a velocity-restricted path planner algorithm named as kinodynamic safety optimal rapidly exploring random tree (KS-RRT
∗
) to plan a near time-optimal path. First, a speed map is introduced to assign different maximum allowable velocity for the sampling points in the workspace, and the traverse time is calculated along the kinodynamic connection of two sampling points. Then the near time-optimal path in the tree-structured search map can be obtained by the rewiring procedures, instead of a distance-optimal path in the original RRT
∗
algorithm. In order to enhance the planner’s performance, goal biasing scheme and fast collision checking technique are adopted in the algorithm. Since the sampling-based methods are sensitive to their parameters, simulation experiments are first conducted to determine the optimal input settings for the specific problem. The effectiveness of the proposed algorithm is validated in several common aircraft parking scenarios. Comparing with standard RRT
∗
and human heuristic driving, KS-RRT
∗
demonstrates a higher success rate, as well as shorter computation and trajectory time. In conclusion, KS-RRT
∗
algorithm is suitable to generate a near time-optimal safe path for autonomous high density parking in semistructured environment.
SDSS-V Algorithms: Fast, Collision-free Trajectory Planning for Heavily Overlapping Robotic Fiber Positioners
