Research on Automated Container Terminal Assignment Problem Based on Multi-agent Architecture Reinforcement Learning Algorithm

In this paper, a reinforcement learning algorithm based on multi-agent architecture is proposed for the task assignment problem of automated guided vehicle (AGV) in automated container terminal (ACT). This algorithm not only takes into account the conditions of AGV decision-making, but also takes into account the difficulty of separating subtasks in other hierarchical reinforcement learning models. The experiment shows that the trained model can effectively shorten the time to complete the same task, and can be used for AGV task assignment problem.