Automated Parking Planning Based on Safety Corridor Constraints and Parking Space Matching

Abstract

The field of Automated Valet Parking (AVP) has garnered significant academic interest in recent years. However, traditional parking planning methods often struggle with maneuvering in compact spaces and large-scale environments. While effective for short-distance parking, these methods frequently encounter increased computational latency and sub-optimal path lengths in long-distance parking scenarios. Fundamentally, the inherent complexity of large-scale facilities extends beyond the practical capacity of conventional single-stage planners, necessitating a multi-stage strategy to balance global search efficiency with local maneuverability. To address these limitations, this paper proposes a hierarchical parking planning framework using a decoupled two-stage approach. By integrating online safety corridor constraints with resolution-adaptive parking space matching, the method combines the global exploration of graph search with the local precision of optimization algorithms. The process employs a two-stage Hybrid A ${}^{\ast }$ strategy: global planning generates an initial trajectory in a low-resolution grid map, subsequently followed by local planning that refines the path to the target point within a high-resolution environment to ensure kinematic feasibility in constrained spaces. Experimental results indicate that this approach enhances algorithm efficiency by approximately 86.58% and reduces path length by an average of 21.68%. The proposed method improves adaptability to compact parking environments and reduces both generation time and path length in large-scale parking lots. Furthermore, the integration of adaptive safety corridor constraints and parking space matching enhances path safety and precision. The effectiveness of the method has been validated through real-world vehicle experiments, demonstrating its feasibility in practical AVP scenarios.