基于故障树的在轨服务目标航天器故障检测

摘要

Knowledge acquisition and representation in constructing on orbit fault detection knowledge base are difficult.To solve this problem, experience knowledge can be used to establish fault tree. Through the qualitative analysis and quantitative analysis of the fault tree, minimum cut sets and importance of the bottom event are obtained, and knowledge redundancy is reduced. The method is demonstrated to be feasible by an example of electricity supply problems of on-orbit spacecraft.%针对构建在轨故障检测知识库中知识获取和知识表示困难的问题,将故障树应用到在轨服务目标航天器故障检测知识库中.采用将专家经验知识建立成故障树的形式,通过对故障树的定性和定量分析,求出其最小割集和底事件重要度,减少知识库中的知识冗余,并以在轨运行的目标航天器上某负载的供电系统发生故障为例进行应用分析.分析结果验证了该方法的可行性.