Simulation-based joint optimization for identifying optimal task allocation strategies to enhance order picking efficiency in robot mobile fulfillment systems

Integrating multiple interrelated operational problems and jointly optimizing them is crucial for enhancing the order-picking efficiency of Robotic Mobile Fulfillment Systems (RMFS). However, most prior studies on RMFS joint optimization have focused on either order-oriented or robot-oriented issues, with limited attention to both aspects simultaneously. Additionally, few studies have comprehensively compared commonly used task allocation strategies and explored their optimal parameter settings. This study proposes a joint optimization framework that integrates order assignment, order sequencing, pod selection, task allocation, pod sequencing, path planning, and conflict resolution, enabling the exploration of optimal task allocation strategies. A novel Traffic Rule-Based Conflict Resolution Strategy (TR-CRS) is developed by refining conflict resolution based on robots’ relative positions. Simulation results show that TR-CRS outperforms the conventional Picking-Status-Based Conflict Resolution Strategy (PS-CRS) in improving overall order picking efficiency. Simulation experiments reveal that the Work Amount-based Task Allocation Strategy (WA-TAS) outperforms other task allocation strategies. However, as the number of robots increases, system efficiency initially rises but then declines, indicating that simply adding more robots does not always improve performance. Further scenario experiments and generalizability assessments indicate that the parameter α significantly affects WA-TAS performance. As order volume increases, both the optimal number of robots and α stabilize, suggesting that using smaller order volumes with sequential α values—requiring minimal simulation time—can determine the optimal parameter configurations. This work offers new perspectives and valuable insights into identifying optimal task allocation strategies and their parameter settings to enhance RMFS efficiency through the emerging simulation-based optimization paradigm.