Scheduling Planning for Multirobot Vehicle Autonomous Transfer in High-Density Storage Yards Based on Dynamic Network Topologies

Rapid intelligent transformation and surging production in the global automotive industry have rendered multirobot autonomous vehicle transfer in high-density yards an urgent challenge. This article tackles the multirobot scheduling planning problem for vehicle transfer across multiple high-density yards by addressing network topology construction, task scheduling, and path planning. First, a dynamic network topology construction method is proposed, complete with criteria for real-time updates. Next, during the scheduling stage, the traditional distance-based calculations are replaced with the A^* search algorithm, and a bidirectional task scheduling strategy is developed to account for potential conflicts in high-density environments. For the planning stage, an enhanced spatio-temporal A^* algorithm based on adaptive conflict detection is introduced, incorporating improvements in node expansion strategies and heuristic functions under turning penalties. Finally, simulation experiments in typical high-density yard scenarios demonstrate that the proposed dynamic network topology reduces the total traveled distance by 10.1% and task completion time by 11.0% compared with conventional structures, while the overall scheduling planning approach further lowers operating costs and boosts efficiency. Furthermore, preliminary verification tests conducted at Yantai Port in Shandong Province further validate the feasibility of our approach.