在串聯系統中, 任一元件失效, 將導致系統無法運作, 有時無法確知造成系統失效的元件, 即為隱蔽資料。本文討論觀測值為區間資料之串聯系統, 其中各元件壽命服從指數分配且其平均壽命與應力間具對數線性關係, 並服從累積曝露模型。考慮型I 設限的階段加速壽命試驗, 分別在元件壽命分配彼此獨立及具Marshall-Olkin 二元指數分配下, 利用期望值-最大概似演算法及遺失資訊法則求得參數之最大概似估計與其費雪訊息矩陣, 進而得到系統可靠度之相關推論。

High reliability products have longer lifetime under normal environment. Accelerated life tests are
usually used to reduce the experiment time. In a series system, the system fails when any of the
components fails, while the cause of system failure may not be observed which is known as masked
data. In this thesis, we consider the step-stress accelerated life tests for series systems with Type-I
censoring, in which the lifetimes of components are exponentially distributed. We not only consider
those distributions are independent, but also consider the Marshall-Olkin bivariate distribution for
two-components series systems. Assume that there exists log-linear relationship between the mean
lifetime of components and the levels of the environmental stress variables under the cumulative exposure model, and the data analyzed are interval data in the sense only the numbers of failures are observed at the times of changing stress levels. Maximum likelihood inference is developed
incorporated with the EM algorithm as well as the missing information principle to achieving the Fisher information. Simulation study is carried out in the reliability analysis. It shows that the proposed method is more accurate and efficient than the bootstrap method.

[1] Bai, D. S., Kim, M. S., and Lee, S. H. (1989). Optimum simple step-stress accelerated life
tests with censoring. IEEE Trans. Reliab., 38, 528–532.
[2] Balakrishnan, N. and Mitra, D. (2012). Left truncated and right censored weibull data and
likelihood inference with an illustration. Computational Statistics and Data Analysis, 56,
4011–4025.
[3] Fan, T. H. and Hsu, T.M. (2011). Statistical inference of a series system with masked data with
correlated log-normal lifetime distributions under type I censoring. International Conference
on Advances in Probability and Statistics-Theory and Applications, December 30, 2011.
[4] Fan, T. H. and Hsu, T. M. (2012). Accelerated life tests of a series system with masked interval
data under exponential lifetime distributions. IEEE Trans. Reliab., 61, 798–808.
[5] Guan, Q. and Tang, Y. (2012). Optimal step-stress test under type I censoring for multivariate
exponential distribution. J. Statist. Plann. Inference, 142, 1908–1923.
[6] Ju, S. K. (2013). Reliability analysis of two-component series system under Marshall-Olkin bi-
variate weibull lifetime distribution. Master Thesis, Graduate Institute of Statistics, National
Central University, Jungli, Taiwan.
[7] Kundo, D. and Dey, A. K. (2009). Estimating the parameters of the Marshall-Olkin bivariate
weibull distribution by EM algorithm. Computational Statistics and Data Analysis, 53, 956–
965.
[8] Li, G. A. (2005). A characterization of the multivariate Marshall-Olkin exponential distribu-
tion and its parameter estimation. Chinese Journal of Engineering Mathematics, 22, 1055–
1062.
[9] Lin, H. H. (2012). Competing risks model of Marshall-Olkin bivariate exponential distribution
under hybrid censoring. Master Thesis, Graduate Institute of Statistics, National Central
University, Jungli, Taiwan.
[10] Louis, T. A. (1982). Finding the observed information matrix when using the EM algorithm.
Methodological , 44, 226–233.
[11] Marshall, A. W. and Olkin, I. (1967). A multivariate exponential distribution. Journal of the
American Statistical Association, 62, 30–44.
[12] Miller, R. and Nelson, W. (1983). Optimum simple step-stress plans for accelerated life testing.
IEEE Trans. Reliab., 32, 59–65.
[13] Miyakawa, M. (1984). Analysis of incomplete data in competing risks model. IEEE Trans.
Reliab., 33, 293–296.
[14] Mukhopadhyay, C. and Basu, A. P. Bayesian analysis of competing risks: k independent
exponentials. Technical report No.516, Department of Statistics, The Ohio State University.
[15] Mukhopadhyay, C. and Basu, A. P. (1997). Bayesian analysis of incomplete time and cause of
failure data. J. Statist. Plann. Inference, 59, 79–100.
[16] Nelson, W. (1980). Accelerated life testing - step-stress models and data analysis. IEEE Trans.
Reliab., 29, 103–108.
[17] Nelson, W. (1990). Accelerated Testing: Statistical Models, Test Plans, and Data Analyses,
John Wiley Sons, New York.
[18] Sarhan, A. M. (2001). Reliability estimation of components from masked system life data.
Reliability Engineering and System Safety, 74, 107–113.
[19] Usher, J. S. (1996). Weibull component reliability prediction in the presence of masked data.
IEEE Trans. Reliab., 45, 229–232.
[20] Usher, J. S. and Hodgson, T. J. (1988). Maximum likelihood analysis of component reliability
using masked system life-test data. IEEE Trans. Reliab., 37, 550–555.