船舶的大型化為目前貨櫃運輸的趨勢，亦使得傳統人工經驗的排艙困難度大幅增加。過去的學術研究皆假設航線上各港口的出口貨櫃需求量為固定，簡化了實際的變動需求，因此本研究針對變動需求下貨櫃船排艙規劃問題，發展一排艙架構，其核心為一個二階段求解程序，內含二個零壹整數規劃模式，以幫助海運業者有效進行其排艙規劃。在第一階段中，本研究將貨櫃船的艙位分區，且將性質相近的貨櫃編為同組，以翻櫃成本最小化為目標，考量船舶平衡、壓櫃及其他相關營運的限制條件，構建一貨櫃組艙區指派模式。在第二階段中，本研究構建一艙位指派模式，依據第一階段的指派結果對每一艙區內的貨櫃進行指派。本研究利用數學規劃軟體發展一演算法進行求解。為測試本研究模式及演算法的實用績效，本研究參考一國籍海運業者的營運資料進行測試，其模式結果與演算效率均頗佳，顯示本研究模式與演算法應可為實務業者之參考。

Recently, container ship capacity has been increased and thus significantly aggravates the difficulty in stowage planning which is usually performed manually in practice. In addition, former studies usually assumed container demand to be constant, that simplifies demand variations in actual operations. Therefore, to solve the problem of container ship stowage plan to meet various demand, in this study we developed a stowage planning framework that contained two 0-1 integer programs embodied in a two-stage process to help naval carriers effectively perform their stowage plans. In the first stage, by assigning container groups with the same OD into the blocks, we developed a contain-block assignment model to minimize the number of container shifting, subject to the ship stability constraint, the container overstow constraint, and the other related operating constraints. In the second stage, a container-slot assignment model was established to assign each container into a slot according to the solution from the fist stage. We used mathematical programming software to develop a solution algorithm. To evaluate the models and the solution algorithm in practice, we conducted case study referring to the data obtained from a Taiwan ocean container liner. The results were positive, showing that models and the solution algorithm could be useful for carriers in practice.