Warehouse Order-Picking Simulation

该数据集基于模拟仓库环境，旨在研究涉及多个移动机器人和人类工人的订单拣选问题。该仿真环境包含了多种配置，用于测试分层多代理强化学习算法。此外，实验过程中使用了WandB进行性能指标的跟踪监测。数据集的规模包括1276个物品位置、16辆自动引导车、8名拣货员，以及每场模拟中的80个订单。任务目标是使用多代理强化学习（MARL）对订单拣选的协调与优化进行研究。

This dataset is built upon a simulated warehouse environment, aiming to investigate the order picking problem involving multiple mobile robots and human workers. This simulated environment provides diverse configurations for testing hierarchical multi-agent reinforcement learning algorithms. Additionally, WandB was utilized to track and monitor performance metrics during the experiments. The dataset encompasses 1276 item locations, 16 automated guided vehicles (AGVs), 8 pickers, and 80 orders per simulation run. The core task objective of this work is to conduct research on the coordination and optimization of order picking using multi-agent reinforcement learning (MARL).