Location assignment for outbound containers with adjusted weight proportion

摘要

This paper studies the location assignment for arriving outbound containers during container-receiving stage. For the problem, the literature assumed that the proportion of the remaining containers on a weight group keeps unchanged throughout the container-receiving process. This assumption is inconsistent with the practice that it should be constantly adjusted according to the containers that have already been received. We therefore propose two other handling ways in this paper, leading to two new dynamic programming models. These two models are compared with the existent model on small-scale instances. For large-scale instances, a two-stage heuristic is proposed. In the first stage, a neighborhood searching heuristic is developed to generate the priority sequence of stacking patterns for each weight group of containers; in the second stage, a rollout-based heuristic is proposed to improve the incumbent solution by simulating more stack alternatives for each arriving container. The numerical experiments show that the model with adjusted weight proportion can significantly reduce state size and improve stacking quality, and that the proposed two-stage heuristic is effective and efficient for large-scale instances.