臺灣博碩士論文加值系統

隨著全球化市場的競爭壓力，許多企業於各地區設有配銷中心及製造廠，競爭的型態不再是企業個體之間的競爭，已經轉變為供應鏈與供應鏈的競爭，所面臨的生產活動，亦從過去的單廠區生產演變成為跨區域的多廠區生產模式。因此，在多廠區生產規劃的環境中，如何在符合產業特性的情況下，有效整合整體供應鏈上採購、生產及運輸規劃便成為重要的議題。
就記憶體模組產業而言，位於產業鏈的中游，在各地區設有配銷中心及製造廠，屬於多階多廠區的生產環境。此外，其原物料記憶體晶片的採購價格會隨著時間的不同而有所差異，且供給數量亦有其限制，因此在進行生產規劃時，必須要考量原物料採購的時間點及其可能造成的存貨成本。此外，由於原物料供給數量有限，當需求大於供給時，如何在有限的供給數量分配至部分需求訂單以達到最高的淨利，亦是記憶體模組產業需考量的重點。另一方面，記憶體模組具備有多階產品替代的特性，相同的完成品可以由多種半成品組裝而成，而相同的半成品又可以由多種原物料製造而成，因此，規劃人員在進行規劃時，必須考量各階品項可供給情況及可能造成的存貨成本，決定以何種原物料及半成品生產而成。
過去的文獻中，甚少針對記憶體模組產業之多廠區生產規劃進行探討，且未同時考量該產業之所有生產特性，因此，本研究提出符合記憶體模組產業的供應網絡生產規劃模式，考慮物料供給限制、原物料採購價格、多階產品替代、各廠區產能限制、生產及運輸前置時間等特性，於多階多廠供應網絡環境中，此模式以有向圖形為基礎之基因演算法求得最高淨利以解決訂單分配的問題。經由實驗結果顯示，本研究方法相較於業界現行作法，可同時考量多筆訂單且決定採購時間點，以達整體之近似最佳解，不論是在總淨利或是缺貨數量皆有明顯的改善。

Due to the competition of globalization market, companies are required to locate distribution centers and manufacturing sites around the world. The type of competition has already changed from individual company to supply chain. Production activity that should be faced has altered from single-plant to current multi-plant as well. Therefore, how to develop an integrated multi-site production planning approach which suits to characteristic of industry to coordinate procurement, production and transportation in supply chain become an important issue in the multi-site production environment.
In terms of memory module industry, which is the part of midstream of industry, companies possess many distribution centers and manufacturing sites that are located in different places geographically. Accordingly, memory module industry belongs to multi-level and multi-site production environment. Besides, the purchasing price and supply quantity of critical material DRAM chip vary over time. Production planners must consider purchasing time of DRAM chip and inventory cost that may cost in production planning. Furthermore, because of limited supply quantity of DRAM chip, how to allocate the finite resources to demand orders for the purpose of largest net profit is another key point for memory module industry. Moreover, in memory module industry, product structure is very complicated since the multiple to multiple substitution relationships. This substitution relationships means same type of finished goods may be assembled by different types of semi-finished goods, and the same type of semi-finished goods may be composed by different types of raw materials. For this reason, production planners should ponder the quantity of each item (including semi-finished goods and raw materials) to decide what types of semi-finished goods and raw materials will be composed to fulfill the orders.
On the current researches, multi-plant production planning for memory module industry or the production planning that deliberate all characteristics of memory module industry have only been considered insufficient. Consequently, the supply network planning model that suits to memory module industry is proposed in this paper, which considers limited supply quantity and purchasing price of critical raw materials, multiple to multiple substitution relationships, limited capacity of manufacturing site, production and transportation lead item. The model is established by directed graph based genetic algorithm that generates the order allocation planning aim for largest net profit. Comparing the SNP model that is proposed in this paper and current method by experiment, the SNP model can consider all demand orders simultaneously and decide the purchasing time of raw materials in order to achieve the global optimization. In the result, significant improvement is shown no matter in total profit or shortage.

摘要 1
ABSTRACT 2
誌謝 3
目　錄 4
圖目錄 6
表目錄 8
第一章 緒論 9
1.1 研究背景 9
1.2 研究動機 11
1.3 研究目的 13
第二章 文獻探討 14
2.1記憶體模組產業介紹 14
2.1.1 記憶體模組產業現況 14
2.1.2 產業鏈關係 15
2.1.3 記憶體模組製程 16
2.1.4 記憶體模組產業特性 17
2.2 供應網絡生產規劃的定義與特性 18
2.3 供應網絡生產規劃相關研究 21
2.4 基因演算法 26
2.4.1 基因演算法介紹 26
2.4.2基因演算法之方法與流程 27
第三章 供應網絡生產規劃模式 30
3.1 記憶體模組產業供應網絡生產規劃現況方法 33
3.2 以有向圖形為基礎之基因演算法 36
3.2.1 以有向圖形(Directed Graph)描述問題 38
3.2.2 基因演算法 48
第四章 實驗設計 81
4.1 實驗方式與環境建構 81
4.1.1 實驗環境 81
4.1.2 實驗因子及基因演算法參數 82
4.2 實驗設計及分析 85
4.2.1 各種環境因子組合之基因演算法參數設定 85
4.2.2 本研究之模式與現況方法比較 89
第五章 結論與未來發展方向 100
5.1 結論 100
5.2 未來發展方向 100
參考文獻 101
附錄一：環境組合B之基因演算法參數設定 103
附錄二：環境組合C之基因演算法參數設定 105
附錄三：環境組合D之基因演算法參數設定 107
附錄四：環境組合E之基因演算法參數設定 109
附錄五：環境組合F之基因演算法參數設定 111
附錄六：環境組合G之基因演算法參數設定 113
附錄七：環境組合H之基因演算法參數設定 115

1.Arntzen, B. C., G. G. Brown, et al. (1995). "Global Supply Chain Management at Digital Equipment Corporation." Interfaces 25: 69-93.
2.Altiparmak, F., M. Gen, et al. (2006). "A Genetic Algorithm Approach for Multi-objective Optimization of Supply Chain Networks." Computers & Industrial Engineering 51: 197-216.
3.Bhatnagar R, C. P., Goyal SK (1993). "Models for multi-plant coordination." Eur J Oper Res 67: 141-160.
4.Chen, S. Y. and C. C. Chern (1999). Shortest Path for A Supply Chain Network. the 4th International Conference, Asia-Pacific Region of Decision Sciences Institute.
5.Chern, C.-C. and J.-S. Hsieh (2007). "A heuristic Algorithm for Master Planning that Satisfies Multiple Objectives." Computers & Operations Research: 3419-3513.
6.DRAMeXchange. http://www.dramexchange.com
7.Goldberg, D. E. (1989). "Genetic Algorithms in Search, Optimization and Machine Learning." Addison-Wesley.
8.Guinet, A. (2001). "Multi-site planning: a transshipment problem." International Journal of Production Economics 74: 21-32.
9.Giglio, D. and R. Minciardi (2003). "Modelling and optimization of multi-site production systems in supply chain networks." IEEE.
10.Holland, J. H. (1975). "Adaptation in Natural and Artifical System."
11.Harland, C. (1997). "Supply chain operational performance roles." Integer Manufacturing System 8(2): 70-78.
12.Ishibuchi, H., N. Yamamoto, et al. (1994). "Genetic algorithms and neighborhood search algorithms for fuzzy flowshop scheduling problems." Fuzzy Sets and Systems 67(1): 81-100.
13.Kanyalkar, A. P. and G. K. Adil (2008). "A robust optimisation model for aggregate and detailed planning of a multi-site procurement-production-distribution system." International Journal of Production Research: 1-22.
14.Lendermann, P., B. P. Gan, et al. (2001). "Distributed Simulation with Incorporated APS Procedures for High-Fidelity Supply Chain Optimization." Winter Sumulation Conference.
15.Lin, J. T. (2007). "A multi-site supply network planning problem considering variable time buckets– A TFT-LCD industry case." Int J Adv Manuf Technol 33.
16.Michalewicz, Z. (1999). "Genetic Algorithms + Data Structures = Evolution Programs."
17.Moon, C., J. Kim, et al. (2002). "Integrated process planning and scheduling with minimizing total tardiness in multi plants supply chain." computers & Industrial Engineering.
18.N., L., X. Xiaofei, et al. (2006). Collaborative Planning in Supply Chains by Lagrangian Relaxation and Genetic Algorithms. The Sixth World Congress, Intelligent Control and Automation: 7258-7262.
19.Kawtummachai, R. and N. V. Hop (2005). "Order allocation in a multiple-supplier environment." Int. J. Production Economics: 231-238.
20.Samxsivan, M. and C. P. Schmidt (2002). "A heuristic procedure for solving multi-plant, multi-item, multi-period capacitated lot-sizing problems." Asia-Pacific Journal of Operational Research 19: 87-105.
21.Shen, W., R. Kremer, et al. (2003). "A collaborative agent-based infrastructure for Internet-enabled collaborative enterprises." International Journal of Production Research 41: 1621-1638.
22.Timpe, C. H. and J. Kallrath (2000). "Optimal Planning In Large Multi-site Production Networks." European Journal of Operational Research 126: 422-435.
23.Vignaux, G. A. and Z. Michalewicz (1991). "A genetic algorithm for the linear transportation problem." IEEE Transactions on System, Man, and Cybernetics 21(3): 445-452.
24.Watson, K. and T. Polito (2003). "Comparison of DRP and TOC financial performance within a multi-product, multi-echelon physical distribution environment." International Journal of Production Research 41: 741-765.
25.Wu, D. (2004). "Multi-Item, Multi-Facility Supply Chain Planning : Models, Complexities, and Algorithms." Computational Optimization and Applications 28: 325-356.