Item Storage Assignment for Mobile-Rack Warehouses in the Industry 4.0 Era
研究了移动货架仓库中物品存储分配问题，通过随机整数规划模型和并行自适应大邻域搜索算法，实现货架移动次数减少2.84%–38.94%，对电商仓储运营有参考价值。
Mobile-rack warehouses have become increasingly popular in online retail due to their efficient order-picking capabilities. They adopt the parts-to-picker mode for order picking, while also introducing a new item storage assignment problem (ISAP). This problem involves determining both the categories and quantities of items assigned to each rack, with the objective of minimizing the expected number of rack movements required to fulfill orders under a given distribution. We formulate ISAP as a stochastic integer programming model and then convert it into a deterministic version using sample average approximation (SAA). To address the vast solution spaces resulting from multidimensional category-quantity correlations, we develop a parallel adaptive large neighborhood search (pALNS) algorithm featuring reduced formulations and customized operators. Experimental results demonstrate a 2.84%–38.94% reduction in rack movements compared to popular methods. Sensitivity analyses reveal that setting 20-24 bins per rack or reserving approximately 20% of rack capacity achieves a favorable balance between productivity and operational flexibility.