透過您的圖書館登入
IP:3.16.70.101
  • 學位論文

基於製程良率之計量型兩計畫驗收抽樣系統

Developing a Two-plan Sampling System for Variables Based on Process Yield

指導教授 : 吳建瑋

摘要


驗收抽樣計畫為品質管制中重要之工具,其可作為買賣雙方間判定貨批是否允收之決策準則。其中一種為「兩計畫驗收抽樣系統」,係由加嚴型與正常型單次驗收抽樣計畫搭配轉換機制而組成,此系統由加嚴檢驗開始,當貨批在加嚴檢驗中連續 次允收即可轉換至正常檢驗,若正常檢驗被拒絕第一次後 次內又被拒絕,則需轉換回加嚴檢驗,此系統較傳統型驗收抽樣計畫更具效率與彈性。 產品品質之衡量,除驗收抽樣計畫之外,運用統計理論分析生產流程穩定度之「製程能力指標」亦是當今廣為應用之方法,其用於評估產品在生產中的製程能力水準以及追蹤與改善產品在生產中的異常因子。其中,雙邊製程能力指標 與單邊製程能力指標 與 (簡稱為 )被業界廣泛使用。 因此,本論文提出計量型之新型態兩計畫驗收抽樣系統,參數為相同樣本數與相異允收臨界值,記為 ,分為兩大部分:第一部分結合指標 ,第二部分則結合指標 ,藉由結合單邊與雙邊製程能力指標以提升本論文之實用價值。此研究除針對各型態之計量型兩計畫驗收抽樣系統進行操作特性曲線與平均抽樣數之分析外,更與傳統驗收抽樣計畫進行比較。最後,藉由操作實際案例,提供使用者在應用此驗收抽樣系統時之程序及準則,以凸顯本研究之貢獻。

並列摘要


Acceptance sampling plans can provide the vendor and the buyer a decision rule for lot sentencing to meet their requirements of product quality. One of them is two-plan sampling system, which consists of single sampling plans with normal and tightened inspections. The system begins from tightened inspection, and switch to normal inspection if lots in a raw are all accepted. Then if there is an additional lot is rejected in the next lots after a rejection under normal inspection, it needs to turn back to tightened inspection. Therefore, it is also called tightened-normal- tightened (TNT) sampling system. Process capability indices can provide measures on the ability of reproducing product units that meet the specifications, and the widely used capability indices are and ( and ). So, several variables single sampling plans have been developed based on process capability indices for controlling the lot or process fraction defective recently. For above reasons, we developed a variables TNT sampling system with new type in this paper. It can be divided into two parts, one is based on and the other one is based on . Combining two kinds of process capability indices can enhance the practical value of this paper. The results of sampling systems indicate that the proposed sampling system is more efficient and flexible than traditional acceptance sampling plans. Lastly, we analyzed a practical case to assist users to learn more from the proposed sampling systems.

參考文獻


1.Aslam M., Wu, C. W., Azam, M. and Jun, C. H. (2016). Switching variable sampling based on the process capability index. Journal of Engineering Research. DOI:10.1080/03610926.2015.1004092.
2.Aslam, M., Yen, C. H., Chang, C. H. and Jun, C. H. (2013). Multiple states repetitive group sampling plans with process loss consideration. Applied Mathematical Modelling, 37(20), 9063-9075.
3.Balamurali, S. and Jun, C. H. (2009). Designing of a variables two-plan system by minimizing the average sample number. Journal of Applied Statistics, 36(10), 1159-1172.
4.Calvin, T. W. (1977). TNT zero acceptance number sampling. American Society for Quality Control Annual Technical Conference Transaction, 24, 35-39.
5.Chou, Y. M. and Owen, D. B. (1989). On the distributions of the estimated process capability indices. Communications in Statistics-Theory and Methods, 18(12), 4549-4560.

延伸閱讀