Incoming Container Schedule-Aware Container Rearrangement Planning based on Reinforcement Learning in Container Terminal

With the increasing volume of maritime traffic, optimizing container placement within terminals has become imperative for enhancing operational efficiency and mitigating congestion. In this paper, we propose a novel container rearrangement planning by considering both the current container stack and the sequence of incoming containers. By modeling the problem as a markov decision process and applying the proximal policy optimization algorithm, we train a reinforcement learning agent to make optimal container rehandling and loading decisions. Numerical results demonstrate that our proposed model outperforms a baseline model that does not consider the upcoming container queue.