论文部分内容阅读
在卫星电源故障诊断问题的研究中,针对系统的知识获取难问题,提出了一种将有向图和CLIPS专家系统相结合的分层诊断方法。这种方法利用有向图知识描述容易和CLIPS推理引擎快速的优势,设计了一个以事实描述系统模型知识的自定义模板结构,同时采用有向图分层策略减小故障搜索空间,然后利用CLIPS提供的快速匹配算法完成诊断推理过程,最后结合已有的专家经验知识给出卫星电源的故障诊断结果。实验结果表明,新方法具有很好的准确性和快速性。
In the research of satellite power failure diagnosis, aiming at the difficult problem of system knowledge acquisition, a hierarchical diagnosis method combining the directed graph and CLIPS expert system is proposed. This method uses the advantage of directed graph knowledge to describe the advantages of easy and CLIPS reasoning engine, designs a custom template structure which factually describes the system model knowledge, and reduces the search space of fault using the hierarchical graph of directed graph, and then uses CLIPS The fast matching algorithm is provided to complete the diagnosis and reasoning process. Finally, the fault diagnosis results of satellite power supply are given according to the existing expert knowledge. Experimental results show that the new method has good accuracy and rapidity.