啟發式演算法應用於記憶體模組廠SMT排程–以K公司為例

Application of Heuristic Algorithm for SMT Scheduling on Memory Module Factory - A Case Study of K Company

傅宏斌(Hung-Pin Fu)

指導教授 ： 陳建良

國立清華大學/其它/智慧製造跨院高階主管碩士在職學位學程/碩士(2021年)

摘要

記憶體模組產業屬於組裝代工，是電子產業供應鏈的最末端，在現代科技與智慧製造的發展一直扮演著非常重要的角色，經過幾十年不斷的進步，供應鏈體系已經相當完整，是屬於成熟的技術。且自2018年中國與美國政府之間開始進行一場持續的貿易戰，再來2019年年末發現新冠狀病毒(COVID-19)，導致各國迫使去做生產基地的調整，台灣藉著中美貿易戰與良好的防疫成果，重建作為科技樞紐的地位，因此台灣製造業的產能需求驟增，工廠端的機台利用率必須更嚴密的管理及提升。
本研究以K公司SMT製程為例，SMT為工廠最昂貴的設備也是整段製程的瓶頸站，所以要如何提高機台利用率，降低在製品庫存，精簡的人力配置，以提升公司的整體營運，是一個重要的課題。此研究主要是針對SMT排程優化以提高機台利用率為主，如何選擇正確的產線機台，銜接相同製成優先以減少換線工時的損失，讓產能最大化進而延伸到後段組裝測試製程的分配平均化，以達到人力安排最佳化。SMT製程排程屬於開放性流程式生產(Open Flow Shop)，在生產中必須同時考慮每一個工作經過機器的順序及每一機器上的工作順序，不同客戶不同產品只能在特定的SMT線生產，更要同時為因應臨時急單需求、機台突發狀況以及機台保養排程。因此，本研究參考相關文獻，以模擬先進規劃排程系統(Advanced Planning and Scheduling System，簡稱APS)於記憶體模組產業SMT生產排程規劃，提出適合此產業的啟發式演算法，比較傳統排程方式，經過實驗並以關鍵績效指標來證明，本研究所提出的方法確實可以有效提升SMT機台利用率、提高訂單達交率，以及縮短後製程平均等待時間程。

關鍵字

記憶體模組產業 ； 先進規劃排程系統 ； 啟發式演算法

並列摘要

Memory module industry belongs to assembly OEM, which is the end of the supply chain of electronic industry. It plays a very important and significant role in the development of modern technology and intelligent manufacturing. After decades of continuous progress, the supply chain system has been evolving thoroughly into a mature technology. Since 2018, there has been a continuous and stalemated trade war between China and the U.S. government, followed by the discovery of new coronavirus (covid-19) at the end of 2019, which has forced countries to adjust their production strategies and change manufacturing bases. With the Sino US trade war and good epidemic prevention achievements, Taiwan has rebuilt its pivotal position to be a science and technology hub. As a result, the production capacity demand of Taiwan's manufacturing industry has increased sharply, the utilization rate of machines at the factory end must be more closely managed and improved.
This study takes the SMT process of K Company as an example. SMT is the most expensive equipment in the factory, and it has been identified to be the main bottleneck of the whole process. Therefore, how to improve the utilization rate of machines, reduce the inventory of work in process, and simplify the workforce allocation to improve the overall operational performance of the company is an important topic. This study is mainly aimed at SMT scheduling optimization to improve machine utilization, how to select the right production line machine, link up the same production priority to reduce the loss of line changing time, maximize the production capacity, and then extend to the equivalent distribution of the later stage assembly and testing process, so as to achieve the optimization of human resources arrangement. SMT process scheduling belongs to open flow shop. In a realistic production environment, the sequence of each work passing through the machine and the sequence of workflow on each machine must be considered simultaneously. Different customers and different product lines can only be produced in a specific SMT line, and scheduling is also capable of accommodating temporary urgent orders, unexpected machine emergencies and machine preventive maintenance at the same time. Therefore, this study, referring to relevant literature, simulates advanced planning and scheduling system (APS) in SMT production scheduling of memory module industry, proposes heuristic algorithm suitable for this industry, benchmarks against traditional scheduling methods, and proves solid key performance indicators through well-defined experiments. The method proposed in this study can effectively increase the utilization rate of SMT machine, improve the order delivery rate, and shorten the average waiting time of the post process.

並列關鍵字

Memory Module Industry ； Advanced Planning and Scheduling ； Heuristic Algorithm

參考文獻

Barker, J. R., & McMahon, G. B. (1985). Scheduling the general job-shop. Management Science, 31(5), 594-598.

Blackstone, J. H., Phillips, D. T., & Hogg, G. L. (1982). A STATE-OF-THE-ART SURVEY OF DISPATCHING RULES FOR MANUFACTURING JOB SHOP OPERATIONS [Article]. International Journal of Production Research, 20(1), 27-45.

Cheng, C.-Y., & Huang, L.-W. (2017). Minimizing total earliness and tardiness through unrelated parallel machine scheduling using distributed release time control. Journal of Manufacturing Systems, 42, 1-10.

Hastings, N., & Yeh, C.-H. (1990). Job oriented production scheduling. European Journal of Operational Research, 47(1), 35-48.

Hastings, N. A., Marshall, P., & Willis, R. J. (1982). Schedule based MRP: An integrated approach to production scheduling and material requirements planning. Journal of the Operational Research Society, 33(11), 1021-1029.