本研究為配銷倉儲空間具有限制下之聯合補貨問題（The Joint Replenishment Problem with Warehouse-Space Restriction），是傳統聯合補貨問題與配銷倉儲具有空間限制下之批量排程問題等兩大主題的結合，其最佳化求解為一個具NP-hard複雜度的問題。在二冪策略的決策情境下，假設產品的需求率、整備成本和存貨持有成本皆為已知常數，對配銷倉儲的管理者而言，其關心的是如何在多產品的補貨系統下，將某些產品合併補貨，以分擔整備成本，降低平均總成本。決策者須決定批量大小、補貨次數及補貨週期，使得平均總成本達到最小，並且調整產品的補貨排程，有效利用倉儲空間，以降低最大的倉儲空間需求，使得排程下的最大倉儲空間需求能滿足倉儲空間限制。換言之，要解決倉儲空間限制下的聯合補貨問題，除了要降低平均總成本，還要確認補貨排程是否為合理可行解。本研究分別利用啟發式演算法與遺傳演算法兩種方式求解。合併運用「批量排程的可行解測試程序」，檢視某組最佳補貨乘數與基本週期是否為合理可行解。依據數據實驗得知，當產品數較小時，啟發式演算法與遺傳演算法，皆可迅速求得一個品質不錯的解。當產品數較大時，啟發式演算法兼具執行時間較短與求解品質較佳的優勢。管理者可依其需求與不同決策情境，選擇適用的求解方式，以此作為補貨批量排程決策的參考。 This study is an extension of the Joint Replenishment Problem (JRP) that takes into accounts the warehouse-space restrictions. The focus of this study is to determine the lot sizes of each product under power-of-two policy to minimize the total costs per unit time and to generate a feasible replenishment schedule of multiple products without exceeding the available warehouse-space. In order to test the feasibility of any given solution, we propose a new heuristic, namely, Proc FT. Also, we propose a search algorithm and a genetic algorithm to secure the optimal solution (i.e., the replenishment frequencies of multiple products and the basic period) in this study. By our numerical experiments, we demonstrate that both solution approaches could effectively solve the JRP with warehouse-space restrictions. Also, for the decision makers who face this problem, we provide the guidelines to take an appropriate solution approach as their decision-support tool under different decision-making scenarios.
