物流中心訂單揀取與分類批次分割方法之研究

Abstract

「訂單別揀取」與「批量揀取」是目前物流中心較常採用的兩種揀貨方式，「批量揀取」的目的是透過加總不同訂單之相同品項合併揀取，減少揀取旅行次數，縮短旅行距離，達到提升揀貨效率的目的。目前「批量揀取」的批次分割方法把求取最短的旅行距離設定為批次分割求解目標，而未將因採批量揀取衍生的分類作業納入考慮，故這樣的批次分割方法，所得的分批結果，當面對因受景氣循環、新零售點加入等因素影響，使得訂單「EIQ」起伏變動時，便無法彰顯「批量揀取」的效益，甚至不若採用「訂單別揀取」。 針對上述的問題，本研究以數學規劃方法，將求取最小揀取與分類工時，設定為批次分割的求解目標方程式，構建同時考慮揀取與分類的批次分割模式，進而發展出「MOTC」與「MOTC+SMD」兩種啟發式求解方法，並選擇某一公司物流中心當案例，蒐集實際資料進行個案研究與敏感度分析，於本研究所設定條件下，這兩種求解方法在個案應用中，及改變包含「EIQ」相關參數值進行的敏感度分析中，都有優於「訂單別揀取」與現有的批次分割方法「SMD」的效率表現。

“Single order picking ” and “ Batch picking ” are two commonly used methods which is widely adopted by most distribution centers nowadays. “ Batch picking” is aim at rising picking efficiency by means of combining identical items in distinct orders and picking them altogether to reduce picking time and minimizing the distance of traveling. Current “Batch picking” typically applies one conventional order batching method which sets minimized traveling distance as the optimal solution of order batching, not taking batch picking-generated sorting works into consideration. Solutions obtained based on such method are not capable of manifest the efficiency- rising capability of “Batch picking” method when factors like business cycle, new retail point setting, which may vary “EIQ” of orders, are taken into account. Under such conditions, current ”Batch picking” sometimes might be even less efficient than “Single order picking”. In order to solve previous mentioned problems, this research applies mathematical non-linear programming method which sets minimized picking time and sorting time as the target solution of order batching equations. Also we constituted order-batching model that take picking and sorting into consideration, and developed two heuristic problem-solving methods called “MOTC” and “MOTC +SMD”. A real distribution center is selected for case study in which relevant data is collected and the results solved by the four methods are compared. A sensitivity analysis is also conducted in this study. It reveals that these two methods outperformed “Single order picking” and “SMD” batch picking in both real world cases and sensitivity analysis of parameters (including EIQ)-varying situations.