基于随机效应模型的可靠性数据转换因子估计

Conversion Factors Estimation for Reliability Data Based on Random Effects Model

融合相似装备在性能试验和作战试验阶段的可靠性试验数据可以获得更为准确的转换因子估计,从而利用转换因子将受试对象的性能试验数据向作战试验折合,以扩充样本量。利用随机效应模型来建模此过程,假设相似装备寿命服从指数分布,转换因子总体服从伽玛分布,通过试验数据估计伽玛分布的参数,进而得出转换因子总体均值估计。在此基础上利用贝叶斯理论,结合各装备本身的试验数据导出各自的转换因子增强点估计和区间估计,实现借助相似装备数据增强受试对象可靠性数据转换因子估计的目标。模拟数据算例分析表明提出的方法具有合理性、可行性及优越性。

A more accurate estimation for the conversion factors can be obtained by fusing the test data of similar equipment in the performance testing(PT)and the operational testing(OT),so that the conversion factors can be used to convert the OT data of the subject to the OT context,and the sample size is expanded.The random effects model is used to model this process.All similar equipment lifetimes are assumed to follow the exponential distribution,and the conversion factors generally follow the gamma distribution.The estimation of the overall mean of conversion factors can be obtained.On such basis,the Bayesian theory is used,the enhanced point estimation and interval estimation of each conversion factor are derived based on the test data of each equipment.The objective of enhancing the estimation of conversion factors with the subject data of similar equipment is realized.The analysis of simulation data example shows that the proposed method is reasonable,feasible and superior.