AI 中文总结
本文针对逐步II型删失下多分量系统的应力强度可靠性问题，构建统一推断框架，结合多种估计方法与贝叶斯分析，经模拟验证后用实际数据应用，并确定最优逐步删失策略。
AI 中文摘要
本文开发了一个统一推断框架，用于解决逐步II型删失下多分量系统的应力强度可靠性问题。采用期望最大化算法得到可靠性的极大似然估计，随后基于缺失值原理确定对应的费希尔信息矩阵和置信区间。为便于对比推断评估，还构建了最大乘积间距估计。在贝叶斯框架内，通过采用有信息和无信息先验模型开展综合分析，利用马尔可夫链蒙特卡罗算法获取合适的汇总结果。通过大量模拟研究分析所有估计量的性能，最后使用一个可靠性数据集展示所提方法的实际应用。此外，采用三种不同的最优性度量确定最优逐步删失策略，并讨论其在可靠性研究中的实用性。
英文摘要
A unified inferential framework is developed to address the stress-strength reliability of multicomponent systems under progressive Type II censoring. The maximum likelihood estimate of reliability is obtained using an expectation-maximization algorithm, followed by the determination of the corresponding Fisher information matrix and confidence intervals based on the missing-value principle. To facilitate a comparative inferential assessment, maximum product spacing estimates are also developed. By employing both informative and non-informative prior models, a comprehensive analysis is conducted within a Bayesian framework, and suitable summaries are obtained using the Markov chain Monte Carlo algorithm. The performance of all the estimators is analyzed through an extensive simulation study. Finally, a practical application of the proposed methodology is presented using a reliability data set. Furthermore, we determine optimal progressive censoring strategies using three different optimality measures and discuss their usefulness in reliability studies.
