應用六標準差手法與TRIZ建構 射出成型製程能力改善與驗證

Research on the Construction of Capability Optimization and Experimental for Injection Molding Manufacturing Process Optimization by using Six Sigma Method and TRIZ

林俊源(Chun-Yuan Lin)

指導教授 ： 周永燦

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2015年)

https://doi.org/10.6840/cycu201500275

摘要

2011年德國舉辦的漢諾威工業博覽會提出「工業4.0 (Industry 4.0)」準則，此準則引領工業走向智慧製造的新世代工業的核心型態。近年來勞力資源短缺，企業逐漸轉型自動化，機檯取代人力需求，而新產品開發到生產上市日程逐漸縮短，必須有效增加生產效率以及提升產品品質，新產品的製程能力流程即成為產業競爭力的重點。現今的製程能力流程在關鍵因子的探討多以工程師的經驗為基準，人為判定失誤導致產品生產失敗風險提高，本研究利用六標準差Define、Measure、Analyze、Improve、Control (DMAIC)之管理步驟、TRIZ方法、實驗設計和倒傳遞類神經預測方法來改善現今的製程能力流程，並透過實證案例塑膠載具進行真平度最佳值之評估。本研究將可應用於縮短智慧工廠之產品生產時間並且穩定產品品質。其結果顯示結合TRIZ、實驗設計及倒傳遞類神經能更加科學篩選出因子，以供實務上做為在提升品質與決策規劃時之參考依據，達到縮短產品的製造時間及預測出產品生產的品質。

關鍵字

製程能力流程、TRIZ理論、倒傳遞類神經、實驗設計、六標準差

並列摘要

Since Industry 4.0 was initially proposed at the Hannover MESSE in Germany in 2011, the industry has transformed toward a new generation - intelligent manufacturing. On account of labor shortage in the recent years, companies have adopted a new business model where labors are replaced by machines, resulting in rapid product and service delivery. Hence, being able to improve productivity and product quality has become the new competitive advantage for companies. And what determines the competitive advantage is its Manufacturing Capability Optimization. However, one of the issues with Manufacturing Capability Optimization nowadays is the dependency on engineers’ experiences which might lead to human errors and exposure to manufacturing failure risks. By looking into product manufacturing reduction for wisdom factory and stabilization of manufacturing quality by performing flatness evaluation on real case - plastic tray, this thesis aims to address the abovementioned issue. Also, it will adopt Six Sigma and DMAIC to further address the flatness problem during the manufacturing.
This thesis defines the features of injection molding manufacturing process of equipments, assesses measurement system and capabilities in manufacturing process, determines critical factors by conducting TRIZ and DOE, and eventually applies to BPN that further refines the existing Predictive Model. The result of this thesis suggests that the combination of TRIZ, DOE, and BPN is a more scientific method to better determine critical factors, and the practical implication contributes to increasing quality and providing a reference for decision making, leading to less time consumed in product manufacturing process and better product quality prediction.

並列關鍵字

Manufacturing Capability Optimization ； TRIZ ； DOE ； Six Sigma ； BPN

參考文獻

李偉正 (2014)，“運用實驗設計和倒傳遞類神經網路結合模式於降低觸控玻璃製程阻抗之研究”，中原大學工業與系統工程所博士論文

曾冠傑(2009) ，“改善手工具品質的穩健設計實證”，中原大學工業與系統工程所碩士論文

陳裕文(2008) ，“田口方法在手機面板熱週期檢測之變形改善”，中原大學工業與系統工程所碩士論文

Ahilan, C., Kumanan, S., Sivakumaran, N., Dhas E., and Raja, J., 2013. “Modeling and prediction of machining quality in CNC turning process using intelligent hybrid decision making tools.”, Applied Soft Computing, 13 (3), pp. 1543–1551

Amir, M. K., Mohsen, A., and Pooneh, S., 2013. “Modeling of TiC-N Thin Film Coating Process on Drills Using Particle Swarm Optimization Algorithm”, Arabian Journal for Science and Engineering June, 38 (6), pp 1565-1571