An Efficient Optimization-Based Framework for Trajectory Planning with Enhanced Hybrid A* and A Modified Travel Corridor Generation Strategy

The combination of Hybrid A* and the optimization-based approach is commonly utilized for trajectory planning in intricate automated parking scenarios. The numerical optimization procedure relies heavily on initial guesses derived from search-/sample-based methods. However, in intricate scenarios with tight passageways, the computational cost of finding an appropriate initial guess may be expensive. To address this issue, this paper presents a two-stage trajectory planner comprised of a coarse trajectory searching stage and a trajectory optimization stage. The primary contribution lies in the coarse trajectory searching stage, wherein a guiding-and-detecting-based Hybrid A* is developed to quickly find the initial path in intricate scenarios. In the coarse trajectory searching stage, the enhanced A* approach and path-trimming strategy are utilized to produce a distance matrix-based guiding path. Subsequently, under the guidance of the path, the guiding-and-detecting-based Hybrid A* is employed to quickly plan a collision-free rough path. A time-optimal velocity profile is then attached to the rough path, turning it into a coarse trajectory. The coarse trajectory serves as the initial guess for the subsequent trajectory optimization stage. The simulation results demonstrate the effectiveness and robustness of our proposed planner.