Joint scheduling of multi-equipment in an automated container terminal under charging-on-the-go mode

Multi-equipment joint scheduling has a significant impact on operational efficiency in automated container terminals. Existing systems often make decisions based on specific conditions, while neglecting dynamic environmental changes. Moreover, the energy consumption and charging processes of automated guided vehicles (AGVs) are frequently overlooked, which are crucial for terminal operations. This paper presents a real-time multi-equipment joint scheduling of quay cranes, yard cranes, and AGVs in automated container terminals under a ‘charging-on-the-go’ constraint. An optimisation model is constructed to minimise the maximum completion time, incorporating operation time constraints, handling sequence constraints, and energy consumption constraints. An improved deep reinforcement learning algorithm is proposed to address this problem in uncertain conditions, featuring comprehensive considerations and strategic designs in state and action spaces. Numerical experiments and comparative analysis verify the effectiveness and superiority of the proposed algorithm. Sensitivity analysis offers insightful guidance for terminal managers in configuring and scheduling operations.