基于Kullback信息融合方法的串联系统可靠性评估

Reliability evaluation of series system based on the fusion of Kullback information

讨论了串联系统在具有多源验前信息的情形下的可靠性评估问题,运用K u llback信息作为分布之间距离的度量,在K u llback信息的融合准则下对多个先验分布进行融合.并以融合后的先验分布作为系统的最终验前分布,对串联系统的可靠性指标进行Bayes估计.最后进行的计算机随机模拟结果表明,文中所提出的方法合理且便于应用.

To discuss the reliability evaluation of series system with multiple prior sources. Using the Kullback information as the distance measure of distribution, and the prior distributions are combined under the fusion rule Of Kullback information. The fused prior distribution is taken as the final prior distribution of the system, and using Bayes method to analyze the reliability performances for the series system, Finally, the computer simulation shows that the estimator in this paper is reasonable and convenient for applying.