摘要
Gamma分布函数Γ(α,β)常常用作贝叶斯指数可靠性增长模型的先验分布密度函数,参数α、β值的不同,将使增长试验的评估结果有很大的差异。引入服从逆Gamma分布的折合因子,得到一种可靠性增长分析的整体推断技术。该技术是闭合的,具有递推分析形式。通过典型示例与其它确定分布参数的方法进行比较,可知这种方法更加接近工程实际。
The Gamma distribution function Γ(α,β) is always used as a prior distribution density function of Bayesian exponential reliability growth model. Different value of α,β will make considerable difference in evaluating results. In this paper, a transform factor which belongs to the inverse Gamma distribution function is introduced and the integrated inferring technology of reliability growth analysis is gained. The simulation results show the method proposed in this paper reasonable.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2004年第4期560-562,共3页
Systems Engineering and Electronics
