AGV“货到人”拣选系统的移动货架选择方法

Rack Selection Method of AGV-Based"Parts-to-picker"Picking System

基于自动导引车(AGV)的“货到人”拣选系统,各品类商品可以拆零存放在不同货架上,同一货架也可以存储多种商品,货架的位置也是随机移动的。面对待拣选的订单,选择哪些货架移动到拣选台,以满足订单的品类和数量要求,并最小化选取移动货架个数,是此类型仓库需要解决的关键问题。针对该问题,文章提出基于订单相似度进行订单分批,再通过线性递减权重的粒子群算法来优化移动货架选择解。由于移动货架的优化问题是NP-hard问题,线性规划求解器难以求解,实验结果表明,与线性规划求解器Lingo相比,文本的两阶段移动货架选择方法是有效的,可以在短时间得出求解方案,大大提升了求解效率。敏感性分析进一步揭示了批次数量、品类数量、货架数量以及订单数量,对搬运货架次数的影响,为管理者提供了决策依据。

In automated guided vehicle(AGV)"parts-to-picker"picking system,each kind of products can be disassembled and stored in a number of movable racks.A rack can also store multiple products,and the positions of the racks are also randomly moved.For a batch of orders to be picked,it is a key problem that needs to be solved in the application of this new warehousing system to determine w hich r acks s hould b e m oved t o t he p icking p latform t o m eet t he p icking r equirements o f p roduct t ypes a nd q uantities and minimize the number of the racks to be moved.This paper proposes an order similarity measure method for order batching,and then optimizes the mobile rack selection solution through particle swarm optimization algorithm with linear decreasing weights.Because the optimization problem of the mobile rack is an NP hard problem,it is difficult for linear programming solvers to solve.The experimental results show that,compared with the linear programming solver Lingo,the two-stage mobile rack selection method proposed in this paper is effective and can obtain a solution in a short time,greatly improving the efficiency of the solution.Sensitivity analyses further reveal the inf luence of batch quantity,category quantity,rack quantity,and order quantity on number of times of rack handling,and provide the important basis of decision for managers.