基于AI-ESTATE标准的贝叶斯网络诊断模型研究

Research on Bayesian Network Diagnostic Model Based on AI-ESTATE Standard

贝叶斯网络在解决复杂装备中不确定性和关联性引起的故障中具有很大优势,建立符合AI-ESTATE标准的贝叶斯网络模型,对实现其通用性至关重要;首先介绍了AI-ESTATE标准的主要内容、体系结构和6种信息模型;其次,在对贝叶斯网络进行分析的基础上,研究如何将公共元素模型扩展为贝叶斯网络模型,详细介绍了拓展过程并给出了拓展框图,同时对各实体和属性进行了说明;最后,针对标准中采用的EXPRESS描述语言编辑工具少、通用性差等问题,采用XML语言对其交换文件进行描述,使得诊断信息具备更好的可移植性和互操作性。

Bayesian Network has great advantages in solving the uncertainty and association fault in complex equipments,setting up the Bayesian Network Diagnostic Model according to AI-ESTATE Standard is of great importance to realize the goal of universality.Firstly,the main contents,architecture and six information models of the AI-ESTATE Standard are introduced.Secondly,extending process from Common Element Model to Bayesian Network Model is analyzed in the paper based on the analysis of Bayesian Network,at the same time,extending figure is given,each entity and attribute are introduced in detail.Finally,aiming at the problem that the EXPRESS language used in the standard has the disadvantages of less editing tools and poor utility,XML is used to describe the exchange document for its universal property,as a result,the diagnostic information are equipped with portability and interplay.