摘 要 人工揀貨作業成本佔整個物流中心總作業成本的15%~20%,因此人工揀貨作業在物流中心之作業中佔極重要的一環。在揀貨區中,揀貨出入口及儲架佈置等硬體設備一旦確定,將會影響後續儲位規劃與揀貨作業路徑規劃之優劣。但大多數的研究均從訂單處理、儲位指派及揀貨路徑這三方面個別處理,較少由揀貨區之設施佈置切入研究。有鑑於此,本研究針對物流中心人工揀貨區進行整體規劃之探討,期望能獲得較佳的揀貨作業規劃,提高人工揀貨區之揀貨效率。在多出入口的觀念下,先設計出入口位置與儲架佈置之10種組合,經由歷史訂單規劃儲位,進而模擬未來訂單規劃揀貨路徑。在此採用遺傳演算法來作為求解之工具,並利用實驗設計來分析遺傳演算法中各參數設定對儲位規劃與揀貨路徑規劃的影響。結果顯示不同的佈置型態與不同的訂單資料,會有不同的儲位指派,進而影響揀貨路徑之規劃。本研究建議唯有以整體的角度來規劃揀貨區,由最初的儲存格擺設、出入口位置、品項分群到儲位指派以及揀貨路徑規劃,探討如何整體規劃揀貨區,避免只侷限單一問題之泥沼,才能真正改善整個揀貨區之作業效率。
ABSTRACT The manual order picking operation cost accounts for 15%~20% of the total operational cost of the distribution center, and hence this operation plays an important role among the various operations for the distribution center. For the order picking area, the design of the I/O location and the layout of the shelves can affect the performance of the consequent storage layout and the picking route planning. Past researches were focused on the solutions of order processing, storage assignment, and picking routing independently. Few of them solved this problem form a facility layout viewpoint. The purpose of this study was to solve the total planning issue for the manual picking area in the distribution center so that the efficiency of the picking operation could be achieved. We considered multiple I/O points and 10 combinations of I/O locations and storage layout by historical data and then simulated the picking routes form future orders. A genetic algorithm was developed to solve this problem and experimental design for the parameters design was conducted to achieve the objective. The results showed that storage layouts and orders combination could result in different storage assignments and picking routes. We suggest that for the planning of an order picking area, the storage layout, I/O point, storage assignment, and picking route design should be considered simultaneously so that the total efficiency of the picking area can be obtained.