基于后验风险的多源先验信息融合方法  

Fusion Method for Prior Distributions from Multiple Sources Based on Posterior Risk

作　　者：陈洪科[1] 

机构地区：[1]厦门大学嘉庚学院,福建漳州363105

出　　处：《装备制造技术》2017年第6期223-225,共3页Equipment Manufacturing Technology

摘　　要：针对多源先验信息Bayes融合中先验分布权重分配问题,提出了一种基于后验风险的权重确定方法——构造后验风险矩阵,并基于后验风险与先验分布权重成反比的原则建立权重求解方程进而求得各先验分布的权重系数。通过算例证明了该方法的有效性。In order to assign justified weight coefficients for prior distributions from multiple sources in Bayes analysis,a method was proposed based on posterior risk. Posterior risk matrix was constructed in this method and then equations for weight coefficients were established based on the principle that weight of prior distribution being inversely proportional to posterior risk. The given numerical example verified the validity of this method.

关 键 词：多源信息融合 后验风险 BAYES 可靠性 

分 类 号：TB114.3[理学—概率论与数理统计]