多阶段可靠性增长试验的增长因子建模与评估

Modeling and Evaluation of Multi-Stage Reliability Growth Test Using Improvement Factor

在多阶段可靠性增长试验中,随着试验阶段数的增加,可靠性水平逐渐提高,相应地现场样本数就逐渐减少,仅利用单个阶段数据进行可靠性评估将带来风险。需要融合历史阶段数据对现阶段可靠性水平进行综合估计。对于失效数据服从Weibull过程的可靠性增长试验,针对截尾时刻的失效强度建立了Bayes分析方法。在多阶段可靠性增长试验中,利用增长因子实现多个阶段试验数据的信息融合与可靠性水平的综合估计。从而获得一种简便易行的可靠性综合评估方法,可以在现场样本较小的情况下获得可信的可靠性评估结果。

In the multi -stage reliability growth tests, the reliability level is gradually promoted when the stages＇ number increases, Because the field samples are decreased, it will bring some risk to use only single stage data to implement the reliability evaluation. So it is necessary to make use of previous stages＇ data to help evaluate the reliability level of the current stage. For the reliability growth test where the failure times follow Weibull process, a Bayesian analysis method is established for the failure intensity at the truncated time. In the multi - stage reliability growth tests, the improvement factor is used to implement the information fusion of multi - stage tests data and the synthetic evaluation of the reliability level. Based on tlTLis simple and convenient Bayesian evaluation framework, the reliability evaluation results can be easily obtained, which are proved to be believable in the case that the field data has small sample size.