一种载人航天概率风险评价数据选用与分析方法

概率风险评价(PRA)方法是一项成熟的量化风险评估技术,而可靠性相关数据不足,制约着该方法在国内载人航天领域的应用。针对这一问题,结合NASA的PRA工作经验,提出了一种适合国内载人航天PRA数据选用与分析的方法,综合运用产品的通用数据与特定数据,采用贝叶斯等数据融合方法,对产品进行量化评价。通过航天典型产品案例,与经典评估方法进行对比分析,结果表明,此方法更能反映产品的失效特性。文章提出的方法,可用于载人航天等具有小子样特征的产品PRA数据的选用与分析,为我国载人航天开展PRA工作提供参考。

Method of Data Selection and Analysis for Human Spaceflight PRA

PRA （probabilistic risk assessment） is a mature risk assessment technology.Because of lack of reliability data,this technology applied in human spaceflight is conditioned in China.In order to solve this problem,a data selection and analysis method of PRA application in human spaceflight is put forward on the foundation of the method of PRA in NASA.Both universal data and specific data of products are used,and the data fusion methods,for example Bayesian,are applied to assess the risk of products.The classical assessment method and the method proposed in this paper are compared with the analysis of representative cases,and the result shows that the latter is more effective.This method can be used in the data selection and analysis in human spaceflight and other less sub-sample products,which provides a reference for conducting PRA in Chinese human spaceflight.