Abstract
In this paper, a two-level search strategy fused with an improved no-fit polygon algorithm and improved bat algorithm is proposed to obtain the layout points of multiple vehicles. Additionally, a space-time scheduling strategy is proposed using the Improved D*Lite Algorithm (ID*Lite) and improved Bezier curve to generate the trajectories of individual vehicles. Furthermore, a conflict resolution strategy is introduced to address the collision conflict problem during multi-vehicle scheduling. The proposed strategies aim to achieve the objectives of shortest scheduling time and smallest array space in order to reduce the time and space cost of vehicle loading on ro-ro carriers. Through a comparison with other algorithms (e.g., at high loading pressure, Scenario 3), the strategies presented in this paper demonstrate advantages such as higher deck utilization (92.234%), shorter path length, fewer collision conflicts, and more balanced weighting. Mainstream methods only focus on a certain part of vehicle transportation, we not only use more stable parking spot generation algorithm and parking order generation algorithm, but also more high-speed multi-vehicle path-finding algorithm, path optimization strategy, multi-vehicle scheduling strategy, which innovatively form a complete vehicle planning process, and at the same time, as an algorithmic framework that can be generalized to any 2d or 3d scenarios.