Trajectory Planning for Automatic Vertical Parking Based on Dynamic Window and Wall Following Approach

Existing methods of traj ectory planning for automatic vertical parking are lack of flexibility.This is due to the fact that the existing mathematical models for traj ectory planning cannot adj ust the paths dynamically because of open-loop and off-line planning.Therefore,a method for traj ectory planning of automatic vertical parking based on dy-namic window and wall following approach is developed.The traj ectory planning problem is first decomposed into a time-free path planning process and a time-related velocity planning process.Then,a wall following strategy is introduced to a-chieve traj ectory planning without a priori traj ectory model,based on an analysis of the Ackermann steering characteris-tics of vehicles.Meanwhile,the wall-following path is used as global heuristic guidance for propelling the dynamic window for speed planning.The feedback and optimization of two processes will dynamically adj ust path planning in each step. The results of a simulation experiment shows that the introduced method can faciliate safe and effective parking,and is capable to make local adjustments dynamically,under an observed accuracy of controlling and location within 0.2 m.