关键词: 调度规则, 自动化码头, 换电式AGV, 蚁群算法

Abstract: As the main equipment for the horizontal transportation of containers in automated terminals, the Automated Guided Vehicle's (AGV) operational efficiency will directly affect the overall efficiency of the terminal. Therefore, the scheduling problem of AGV has gradually become one of the research hotspots of terminal optimization problems. At present, the research on the AGV scheduling problem in automated terminals lacks consideration of the charging or battery swapping process, and few scheduling decisions are made in the synchronous loading and unloading operation mode.
Given the operation plan of the quay crane and yard crane, this paper considers the characteristics of synchronous loading and unloading operation mode and generates the AGV schedules to realize the joint decision of its container operation sequence and battery swapping time. The AGV scheduling method proposed in this paper has the following two purposes: One is to maximize the operational efficiency of the terminal, that is, to reduce the waiting time and delay time of container tasks; The other is to minimize the energy consumption of AGV, that is, to reduce the proportion of no-load and waiting time of AGV.
Therefore, this paper takes the minimization of total time cost as the scheduling optimization goal, and comprehensively considers four specific indicators such as no-load time, AGV waiting time, task waiting time, and shore delay time to balance AGV energy consumption and the overall efficiency of the terminal. Aiming at the above joint decision-making problem of AGV operation and battery swapping, an integer programming model is established. According to the characteristics of synchronous loading and unloading operation mode, a heuristic rule based on “task relative priority” is designed. At the same time, in order to reduce the mutual waiting time between AGV and task, a scheduling rule of “earliest available time” is introduced, and the node transfer rule and pheromone update process of the ant colony algorithm are improved accordingly. In order to verify the effectiveness of the proposed model and method, a numerical example is designed based on the data of the fourth phase of Yangshan Port, and different rules and methods are compared and analyzed. The instance analysis results show that in the synchronous loading and unloading mode of the automated terminal, the “relative priority of tasks” scheduling rule proposed in this paper reduces the idle time of AGVs, improves the utilization rate, and reduces energy consumption. At the same time, the waiting time between AGV and tasks is also reduced, and the satisfaction rate of quay cranes and yard cranes is increased, which improves the overall operation efficiency of the automated terminal. The improved ant colony algorithm can also obtain better solutions and improve the search efficiency of the algorithm while ensuring the convergence speed. In addition, the proposed scheduling method can be further extended to other application scenarios of battery-swapping AGVs.
However, there are still some shortcomings in this paper. The model allocates container tasks under a static environment and does not consider the dynamic changes and uncertain factors of the actual production environment, such as the random arrival of tasks. How to take the uncertainties into account in the AGV scheduling process to generate a robust schedule to make a better response in dynamic environments will be a future research direction.

Key words: scheduling rule, automated terminals, power-changing AGV, ant colony