摘要
为降低瞬时辐照抽样检验中现行经典非参数法的评价结果的保守性,引入样本空间排序法,从理论和实例两方面计算无失效数据的生存概率置信下限,并与经典非参数法进行对比。结果是,样本空间排序法不仅提高生存概率的置信下限,还增大这些置信限所对应的剂量率范围。该研究结果表明样本空间排序法提高了数据利用率,降低了保守性,可减少试验成本。
The non-parametric method is conventionally used to evaluate data in lot acceptance testing （LAT） of transient i- onizing radiation. But the evaluation results are very conservative and will lead to an increase of the trial cost. The ordering meth- od in the sample space was introduced and applied to compute the lower confidence limit of survival probability based on zero-fail- ure data and was compared with the non-parametric method both theoretically and via a practical LAT on QG- I. It is concluded that the ordering method can not only improve the lower confidence limit of survival probability but also expand the scope of dose rate corresponding to the lower confidence limit. Results indicate that the ordering method improves data utilization and makes the evaluation less conservative and can save the trial cost.
出处
《强激光与粒子束》
EI
CAS
CSCD
北大核心
2013年第10期2753-2756,共4页
High Power Laser and Particle Beams
关键词
瞬时辐照
无失效数据
区间删失数据
样本空间排序法
transient ionizing radiation
zero-failure data
interval censored data
ordering method in sample space
