本計劃將探討串聯系統之加速壽命試驗的相關可靠度分析，並假設系統中個物件之壽命分布為廣義伽瑪分布。在串聯系統中，任一物件損壞則導致系統失效，而在實務中時常無法得知是何物件造成系統失效，亦即觀測之資料為隱蔽資料。當隱蔽資料比例太高時，資料本身訊息不足，使得傳統的最大概似方法有失精確，而貝氏統計則是值得考慮的另一種方法。本計劃將分別探討最大概似方法與貝氏方法之可靠度分析。另一方面，一般常用的壽命分布，如指數分布、韋伯分布、對數常態分布等皆為廣義伽瑪分布之特例，本計劃並將討論資料之適合性問題。 In this proposal, we will discuss the reliability analysis of a series system under accelerated life tests when the components are assumed to have independent generalized gamma lifetime distributions. In a series system, the system fails if any of the components fails. It is often to include masked data in which the component that causes failure of the system is not observed. When the proportion of the masking data is high, ML approach fails due to lack of information. Bayesian approach is an appropriate alternative in such case. Hence, we will also study the Bayesian approach incorporated with subjective prior distribution with the aid of MCMC method. On the other hand, commonly used life distributions such as the exponential distribution, Weibull distribution, and lognormal distributions are all special cases of the generalized gamma distributions. In this proposal, we will also discuss the problem of the goodness of fit. 研究期間：10008 ~ 10107
