The research focuses on the scheduling optimization of Automated Guided Vehicles (AGVs) in intelligent logistics systems within process manufacturing. Addressing issues in traditional AGV scheduling strategies, such as low handling efficiency and path conflicts in multi-vehicle and complex path and warehouse environments, a hierarchical intelligent scheduling method for multi-AGVs based on Bidirectional A* is proposed. Firstly, a multi-AGV hierarchical scheduling strategy is designed, dividing tasks into workshop-level and warehouse-level scheduling to enhance efficiency and task execution priority. Secondly, an intelligent scheduling algorithm combining Particle Swarm Optimization is introduced to achieve smart AGV scheduling, minimizing path conflicts, optimizing path planning, and improving overall system operational efficiency. Finally, through case study simulations, the proposed hierarchical intelligent scheduling strategy is validated to significantly enhance production efficiency, reduce energy consumption, and shorten production cycles, thereby offering an effective approach for optimizing multi-AGV scheduling in intelligent logistics systems.