科技日新月異，傳統的壽命測試已經不適用於高可靠度產品，因此透過加速壽命試驗可以較為迅速得到產品的可靠度資訊，利用較為嚴格的環境測試條件讓產品提早失效，以估計產品在正常環境下的產品壽命；為了降低成本跟供給需求，現在工廠大多都採取大量生產的方式來降低成本，但是在準備出廠產品時，必須要保證產品品質，若採取全部檢驗的話將是費時費工的工程，因此若採取驗收抽樣的方式，透過抽取一定比例樣本去推估對總體產品的認識，將會是高效率並有用的方法；近年來已有許多學者建立加速壽命試驗下的抽樣計畫，利用抽樣後的樣本，採取加速壽命試驗檢測產品的可靠度，建構允收標準檢定統計量，最終決定此批貨品是否允收，本研究將透過統計推論決定抽樣樣本、以及加速應力值應該要設置為多少才能達成最小的抽樣變異。
Weibull分配常用來當作可靠性分析和壽命試驗的理論基礎，因此本研究假設產品壽命服從於Weibull分配，而過去文獻上加速壽命試驗抽樣計劃只有探討兩個應力水準，但是過往相關研究都是假設模型的參數與應力是線性的，兩個加速應力水準無法驗證是否符合線性假設，過往也有學者研究發現在兩個應力水準下的加速壽命試驗比三個應力水準下的加速壽命試驗較不穩健，因此本研究引用Meeker and Hahn (1985)提出的4:2:1理論應用在本研究上，驗證在加速壽命試驗的抽樣計畫上，三個應力水準會比兩個應力水準來得穩健，透過滿足消費者風險與生產者風險、最小化允收標準檢定統計量的漸進變異，建構一個在三應力水準下的加速壽命試驗抽樣計畫，並且與傳統的二應力水準下的加速壽命試驗抽樣計畫做比較。

關鍵字 : 加速壽命試驗、4:2:1比例方法、抽樣計畫、Weibull分配

In the past, accelerated life test sampling plans (ALTSPs) in the literature only discussed the two-accelerated stress level, but previous related papers have assumed that the parameters of the model and the stress are linear relationship, and the two-accelerated stress level cannot be verified whether they conform to the linear assumption, and some authors have found that the accelerated life test (ALT) under two-accelerated stress level is relatively less robust than the three-accelerated stress level. In this paper, the 4:2:1 method proposed by Meeker and Hahn (1985) is applied to this paper, which verify the sampling plan in the ALT, the three-accelerated stress level also be more robust than the two-accelerated stress level.
This paper determines the number of samples and the value of the accelerated stress by satisfying consumer risk and producer risk, minimizing the asymptotic variance of test statistic for deciding the lot acceptability are obtained. Finally, this paper designs ALTSPs under three-accelerated stress level and compare it with the traditional ALTSP under two-accelerated stress level.

Key words: accelerated life test, 4:2:1 method, sampling plan, Weibull distribution

[1] Attia, A. F., Aly, H. M., & Bleed, S. O. (2011). Estimating and planning accelerated life test using constant stress for generalized logistic distribution under type-I censoring. ISRN Applied Mathematics, 2011, 203618.
[2] Ding, C., Yang, C. Y., & Tse, S. K. (2010). Accelerated life test sampling plans for the Weibull distribution under type I progressive interval censoring with random removals. Journal of Statistical Computation and Simulation, 80(8), 903-914.
[3] Fan, T. H., & Yu, C. H. (2013). Statistical inference on constant stress accelerated life tests under generalized gamma lifetime distributions. Quality and Reliability Engineering International, 29(5), 631-638.
[4] Guan, Q., Tang, Y. C., Fu, J. Y., & Xu, A. C. (2014). Optimal multiple constant-stress accelerated life tests for generalized exponential distribution. Communications in Statistics-Simulation and Computation, 43(8), 1852-1865.
[5] Kantam, R. R. L., Rosaiah, K., & Rao, G. S. (2001). Acceptance sampling based on life tests: log-logistic model. Journal of Applied Statistics, 28(1), 121-128.
[6] Leemis, L. M. (1986). Lifetime distribution identities. IEEE Transactions on Reliability, 35(2), 170-174.
[7] Li, X., Chen, W., Sun, F., Liao, H., Kang, R., & Li, R. (2018). Bayesian accelerated acceptance sampling plans for a lognormal lifetime distribution under type-I censoring. Reliability Engineering & System Safety, 171, 78-86.
[8] Li, X., Gao, P., & Sun, F. (2015). Acceptance sampling plan of accelerated life testing for lognormal distribution under time-censoring. Chinese Journal of Aeronautics (Vol. 28, pp. 814-821).
[9] Ma, H., & Meeker, W. (2008). Optimum step-stress accelerated life test Plans for log-location-scale distributions. Naval Research Logistics (NRL), 55, 551-562.
[10] Meeker, W. Q. (1984). A comparison of accelerated life test plans for Weibull and lognormal distributions and type I censoring. Technometrics, 26(2), 157-171.
[11] Meeker, W. Q., & Hahn, G. J. (1985). How to plan an accelerated life test: Some practical guidelines. ASQC Quality Press.
[12] Meeker, W. Q., & Nelson, W. (1975). Optimum accelerated life-tests for the Weibull and extreme value distributions. IEEE Transactions on Reliability, R-24(5), 321-332.
[13] Meeter, C. A., & Meeker, W. Q. (1994). Optimum accelerated life tests wth a nonconstant scale parameter. Technometrics, 36(1), 71-83.
[14] Montgomery, D. C. (2007). Introduction to statistical quality control. John Wiley & Sons.
[15] Pascual, F. (2006). Accelerated Life Test Plans Robust to Misspecification of the Stress—Life Relationship. Technometrics, 48, 11-25.
[16] Pascual, F., & Montepiedra, G. (2003). Model-Robust Test Plans With Applications in Accelerated Life Testing. Technometrics, 45, 47-57.
[17] Santos-Fernandez, E., Govindaraju, K., & Jones, G. (2014). A new variables acceptance sampling plan for food safety. Food Control, 44, 249-257.
[18] Schneider, H. (1989). Failure-censored variables-sampling plans for lognormal and Weibull distributions. Technometrics, 31(2), 199-206.
[19] Seo, J., Jung, M., & Kim, C. (2009). Design of accelerated life test sampling plans with a nonconstant shape parameter. European Journal of Operational Research, 197(2), 659-666.
[20] Tang, L.-C. (1999). Planning for accelerated life tests. International Journal of Reliability, Quality and Safety Engineering, 6(03), 265-275.
[21] Tang, L.-C., Tan, A.-P., & Ong, S.-H. (2002). Planning accelerated life tests with three constant stress levels. Computers & industrial engineering, 42(2-4), 439-446.
[22] Tang, L.-C., & Yang, G. (2002). Planning multiple levels constant stress accelerated life tests. Annual Reliability and Maintainability Symposium. 2002 Proceedings (Cat. No. 02CH37318).
[23] Watson, G., & Wells, W. (1961). On the possibility of improving the mean useful life of items by eliminating those with short lives. Technometrics, 3(2), 281-298.
[24] Wu, S. J., Hsu, C. C., & Huang, S. R. (2020). Optimal designs and reliability sampling plans for one-shot devices with cost considerations. Reliability Engineering & System Safety, 197, 8.
[25] Yang, G.-B. (1994). Optimum constant-stress accelerated life-test plans. IEEE Transactions on Reliability, 43(4), 575-581.
[26] Yang, G.-B., & Ling, J. (1994). Best compromise test plans for Weibull distributions with different censoring times. Quality and Reliability Engineering International, 10(5), 411-415.
[27] Zhang, J., Liu, F., Liu, Y., Wu, H., Wu, W., & Zhou, A. (2012). A Study of Accelerated Life Test of White OLED Based on Maximum Likelihood Estimation Using Lognormal Distribution. IEEE Transactions on Electron Devices, 59, 3401-3404.