摘要
物流中心的電子標籤輔助揀貨系統可分成「摘取式(Pick-to-light)」與「播種式(Put-to-light)」兩種，所謂「摘取式揀貨系統」即貨架上安裝的標籤所對應的是商品，揀貨人員依標籤的指示，自貨架上將商品取下以滿足手上的客戶訂單；至於「播種式揀貨系統」跟「摘取式揀貨系統」正好相反，貨架上安裝的標籤所對應的是客戶訂單，揀貨人員將批次匯總後之商品，依標籤的指示，並分發至訂購客戶架上。 本研究主要針對「播種式揀貨系統」內部訂單揀貨作業為主要探討重點，針對多橫向走道（cross-aisle）環境，配合吾人提出的揀貨方式模擬出不同的結果，本研究主要可分為四個階段：第一階段為輸入訂單之集中型態、第二階段為品項分群，包括使用先前學者提出的相似係數以及群集分析法、第三階段為備料區品項指派、最後一階段為訂單指派儲位順序，前一階段必須完成後才做下一個階段；本研究即是用以3種不同的訂單型態水準、4種相似係數水準、3種聚群法則水準、2種備料區指派水準，及13種訂單儲位配置水準，共有936種組合進行實驗，並以揀貨作業總行走距離為績效，求取最佳效果組合。結果發現訂單愈呈集中型態，吾人所提出之因子法則對節省行走距離愈有效，在此訂單型態下相似係數計算法、品項聚群法則及備料區指派法則有顯著影響，但在品項平均需求型態或品項需求平均數大的訂單型態下，相似係數計算法、品項聚群法則及訂單指派準則有顯著差異，備料區指派法則反而沒有顯著差異。相似係數計算法、品項聚群法則、備料區指派法則與訂單指派準則各因子的最佳水準分別為SC1、CA3、AL2與OC11。
並列摘要
According to the picking operation in Distribution Centers, the Computer Aided Picking Systems could be divided into “Pick-to-light” and “Put-to-light”. In the “Pick-to-light” system, the digital tags on racks correspond to items and indicate pickers to pick items. In the other hand, the tags correspond to customers’ orders in the “Put-to-light” system. Pickers batch the necessary items of these orders and then assort the needed items to orders by the indication of tags. The study is based on the picking operation of “Put-to-light” systems and experimented with simulation by the proposed methods under multiple cross-aisles warehouses environment. There are four phases in the study: the first phase is types of input orders, the second phase is item grouping, including the similar coefficients and the cluster analysis which proposed in previous research, the third phase is assignment of assorting zones, and the last is sequence of the orders that have be allocated to bins. Once the previous one phase has finished, the next phase will begin. The study is experimented with three different order type levels, four similar coefficient levels, three cluster rule levels, two assignment of assorting zone levels, and thirteen order allocation criterion levels. There are totally 936 combinations and the pickers’ total traveling distance is used as performance measurement to measure every combination. The result is as concentrated as consisting of orders, the proposed methods are more efficacy in saving traveling distance. When the system under the concentrated order type level, the similar coefficient, cluster rule and assignment of assorting zone factors are significant. But when it under uniform order type level, the similar coefficient, cluster rule and order allocation criterion factors are significant. Finally, the best level of similar coefficient, cluster rule, assignment of assorting zone and order allocation criterion are SC1, CA3, AL2 and OC11.