逐步增加Ⅱ型截尾样本下k/N(G)系统可靠性指标的置信区间

Confidence Interval of the Reliability Performances for k/N(G) System Based on Progressively Type-Ⅱ Censored Samples

在逐步增加Ⅰ型截尾样本下,研究k/N(G)系统可靠性指标的区间估计问题。假设部件寿命服从指数分布,利用极大似然法和Bayes方法,首先给出了部件失效率的Bayes近似置信区间,其次推导出了系统可靠度和平均寿命的Bayes近似置信区间的计算公式。最后给出随机模拟例子,并对置信区间的精度进行了讨论。

Based on progressively type-Ⅱcensored samples,confidence interval of the reliability performances for k/N（G） system are investigated.Suppose that the lives of all units in the system are independent and have identical distribution.Firstly,Bayesian approximate confidence interval of failure rate is obtained by using Bayesian and maximum likelihood approaches.Next,the expressions for calculating Bayesian confidence interval of the reliability function and average life are also presented,and an illustrative example is examined numerically by means of the Monte-Carlo simulation.Finally,the accuracy of confidence interval is discussed.