由於網購市場的蓬勃發展以及便利商店快速展店的影響,實體量販店近年來紛紛提供網路訂購快速配送的服務,轉型為一個結合實體與虛擬的(Click-and-Mortar)多通路零售組織。然而,量販業者往往為節省成本,僅運用原有的賣場作為發貨中心,未針對揀貨作業進行妥善的規劃、欠缺有效率的方法來進行訂單處理,也無適當的揀貨路徑規劃,導致揀貨作業的總移動距離過長,而耗費大量的時間在揀貨而使得賣場無法快速完成網路訂單,也致使人工與作業成本居高不下而影響獲利。 有鑑於此,本研究提出一個有效機制解決量販賣場網路訂購快速配送模式下訂單批次化與最短路徑規劃問題。此機制分為兩個階段,第一階段,將大量的網路訂單運用位階順序分群演算法(Rank Order Clustering, ROC)進行適當的批次化(Batching)處理,以利於多張訂單能同時作業;第二階段,利用基因演算法(Genetic Algorithm, GA)進行最短路徑規劃,並以視覺化方式呈現揀貨路徑,解決揀貨人員在商品揀貨路徑上不必要的行走。另外,本研究也探討當賣場進行商品熱賣區的規劃時,是否能縮減揀貨人員在商品揀貨時可能的移動範圍。 本研究透過窮舉法(The Method of Exhaustion)與基因演算法的結果比較,結果顯示基因演算法的運算結果具有正確性。此外,本研究進行兩個實驗,其結果顯示本研究所提出的批次化方法能有效縮短揀貨距離,再者,隨著熱賣區放置商品的增加,亦能達到縮短揀貨距離的效果,因此,驗證本研究提出之兩階段機制確實有效。
With the rapid development of online shopping market, some hypermarkets provide online ordering and express delivery services to transform their physical organizations into click-and-mortar ones, in order to make more profits. However, this model did not work very well because of a lack of appropriate order batching and inefficient picking operations, which result in too long travel distances for picking. To cope with the above-mentioned problem, this thesis presents a two-stage mechanism to find the solutions. At the first stage, online orders are clustered and batched using a rank order clustering (ROC) algorithm. At the second stage, genetic algorithm (GA) is used to find the shortest picking route and their results are displayed by a visualized chart. In addition, this study investigates the influences of setting up a hot-sale zone on the picking. To validate the correctness of the GA program, the method of exhaustion (MOE), which can obtain optimal solutions, is employed and its results are compared with those from GA. The results reveal that GA can obtain optimal solutions in small-sized cases and is time-efficient. In addition, a number of experiments were performed and the results show that the proposed two-stage mechanism is effective and the ROC algorithm can significantly shorten the picking distances. Moreover, the setup of a hot-sale zone influences the picking distance positively.