二階段供應鏈之協同排程

Abstract

在這篇論文中，我們將探討Concurrent open shops中協同合作的可能，並且不單只是考慮最小化完工時間總合，更加入權重的概念，做為最小化的目標式。因為不單單只是完工時間會對成本造成影響，應該要加入權重來詮釋operations或orders 之間的相對重要性，以及反映現實市場中，可能存在買方市場或賣方市場這種買賣雙方地位不平等的情況。我們分別依據位置變數及相對位置變數提出兩個整數規劃模式來闡述這個題目。接著，我們提出一個由WSPT演算法、NEH演算法與variable neighborhood search所構成的三階段近似解演算法，可以在短時間內為此計算複雜度極高的最佳化問題求得不錯的近似解。最後，我們進行實驗模擬並根據實驗結果，從多面向分析此三階段演算法的求解效益與效率。

In this thesis, we investigate a scheduling problem in concurrent open shops incorporating coordination among all participants. Weights are assigned to not only orders (jobs) but also product items (operations) to reflect the fact that any operations, parts or items need to be included in the formation of objective functions and the fact that the positions of buyers and sellers may not be equal in real-world application, such as buyers' market or sellers' market. Two integer programming formulations, based on positional variables and sequencing variables, of the studied problem are presented first. We then propose a three-phase algorithm, which comprises of a WSPT heuristic, the NEH algorithm and the variable neighborhood search, to produce approximate solutions to the computationally intractable problem. Finally, we will analyze the performance of the three-phase algorithm through the computational results.