A Multi-Objective Parking Space Decision-Making and Path Planning Method

Autonomous decision-making and path planning in automated valet parking can improve the safety and intelligence of vehicles. As the complexity of executing tasks continues to increase, autonomous vehicles need to perform decision-making with multiple variables and target points. To plan a path in such cases, balancing computation efficiency and search range remains challenging. To address the problems mentioned above, this paper proposes a multi-objective parking space decision-making and path planning method. First, considering the uncertainty of parking occupancy distribution, a multi-objective function is designed to select optimal parking space with multiple midpoints. Then, based on the target parking space and sequence of multiple midpoints in the decision-making stage, the path planner generates a feasible path by using the prior information of the parking lot's topology. The experiment demonstrates that the proposed method achieves a significant improvement in efficiency and cost compared with traditional algorithms.