運用加權式的關聯規則進行產品儲位規劃

Abstract

倉儲管理中儲位規劃講求的是降低揀貨行走路徑、減少人力成本。早期儲位規劃將庫存做產品ABC分類，將同是高價值或高週轉率的產品放置同一區域並接近進出點來減少揀貨行走距離；或利用EIQ分析（Order Entry, Item, Quantity）將客戶的訂單及購買的品項與數量進行交叉分析，找出極可能一起出貨及出貨頻率高的產品，將其擺放至相鄰區域達到揀貨距離縮短。傳統的作法著重在過往資料的分析，而本研究以資料探勘技術，尋找出貨產品間的關聯規則，並以求得的增益值及產品預計出貨量為權重，建立產品儲位指派先後順序之模型。不僅藉由使用預計出貨量讓規劃更具意義，也將研究過程公式化。最後透過個案公司進行實例分析，研究結果顯示，使用本研究之方法所降低的行走路徑為原始個案公司儲位配置方法的22%，同時統計結果顯示新舊方法存在顯著差異，本研究方法可有效降低揀貨之行走距離。

The objective of storage assignment in the warehouse management is to reduce the order picking distance and save labor cost. Most literature related to storage assignment policies focus on the class-based policy and EIQ (Order Entry, Item, Quantity) analysis to find the product properties, such as turnover rate and order frequency. The traditional approaches used the analysis of the transaction data, while this paper applies the data mining to find the weighted association rules with the product forecast ship quantity. The weighted association rule emphasizes the relationship of products and turnover rate in future. Finally, the real world case shows that the weighted association rule decreases travel distances by 22%, and the statistical results show that our proposed mechanism can significantly improve the performance of reduce the picking distances.