摘要
事件树方法是安全风险分析的基本技术，其分析结果的可信度在很大程度上取决于控制系统失效概率估计的精度。由于可靠性试验费用昂贵，试验次数有限，因此控制系统失效概率的估计往往较粗糙。提出的贝叶斯分析方法能够有效地利用系统运行过程的事故统计数据，对控制系统失效概率进行修正，可显著提高估计的精度。
Event tree method is the basic technology of safety risk analysis, but the credibility of analysis mainly depends on the precision of failure probability parameters'o estimate of control systems. For the too high expense, the number of reliability experiment is limited, so the estimate of the parameters of failure probability is rough. This paper introduces a method of Bayesian analysis, which can update the estimate of the parameters of failure probability based on the data during the system's operation and improves the precision of estimate greatly.
出处
《系统工程与电子技术》
EI
CSCD
1999年第9期78-80,共3页
Systems Engineering and Electronics
