Abstract: To improve the picking efficiency of Robotic Mobile Fulfillment System (RMFS) order picking system in manufacturing enterprises,the impact of replenishment on the picking efficiency in the order picking process was analyzed,and the storage allocation problem under real-time replenishment was studied.Aiming at the shortest total handling distance in the two stages of replenishment and picking,an integer nonlinear programming model was established.A storage allocation algorithm based on binary network and an improved gray wolf optimization algorithm was proposed.The first two algorithms were used to effectively solve the storage allocation problem of RMFs order picking system based on real-time replenishment.Experiments showed that the designed storage allocation algorithm and improved gray wolf algorithm had obvious advantages in solution accuracy and stability compared with genetic algorithm,traditional gray wolf algorithm,improved artificial bee colony algorithm and improved gray wolf algorithm with Lévy flight,and effectively improved the operation efficiency of RMFS order picking system under different warehouse sizes and order picking sizes.

Key words: real time replenishment, robotic moble fulfillment system, storage allocation, gray wolf optimization algorithm

摘要： 为提高制造型企业基于移动机器人的拣货系统（RMFS）的拣选效率,分析订单拣选过程中补货对拣选效率的影响,对其实时补货情况下的储位分配问题进行研究,以补货和拣货两阶段总搬运距离最短为目标,建立整数非线性规划模型,提出基于二分网络的储位分配算法和改进灰狼优化算法,利用前两个算法有效解决了基于实时补货情况下的RMFS订单拣选系统储位分配问题。实验表明,设计的储位分配算法和改进灰狼算法与遗传算法、传统灰狼算法、改进人工蜂群算法、引入Lévy飞行的改进灰狼算法相比,在求解精度和求解稳定性上有较明显的优势,在不同仓库规模和订单拣选规模下有效提高了RMFS的作业效率。

关键词: 实时补货, 基于移动机器人的拣货系统, 储位分配, 灰狼优化算法