Bayes方法应用中验前分布选择的一致性与可信性原则

Bayes方法已在装备的参数估计与决策分析中得到了广泛应用。现有的研究主要集中于验前分布的构造与方法的工程应用，而对验前信息之间的一致性、验前信息的可信性研究较少。给出了验前信息一致性的定义与检验方法，建立了验前可信性分析的假设检验模型。以此为基础，建立了基于一致性与可信性的验前分布融合模型。最后，以装备可靠性分析为例，对一致性、可信性的计算过程进行说明。

Principles of consistency and reliability for prior distributions selection in Bayesian method application

Bayesian method has been widely used in equipment parameter evaluation and decision making. However, existing studies focus on the structure of prior distribution and the engineering applications of the method, and there are few discourses on the analysis of consistency among prior distributions and the reliability of prior informations. The definition and testing methods for consistency analysis are given, the hypothesis testing model of prior reliability is built. Based on an analysis of consistency and prior reliability, a prior information fusion model is constructed. For an equipment reliability analysis example, the calculation process of consistency and prior reliability is exhibited.