摘要
阐述了故障诊断推理过程中知识的概念、类别和关系,从知识的获取、组织和表示等方面对定性和定量知识的集成进行了研究,并提出了基于定性推理和定量仿真集成的综合诊断推理方法,结合复杂系统试车数据的仿真诊断结果表明,该方法能有效地消除推理结果中的奇异性,从而改善了定性推理方法的性能。
The concept, kind and relation of knowledge in fault diagnosis are described firstly, then the integration of qualitative and quantitative knowledge is discussed in the aspects of acquisition, management and expression. Secondly, an integrated diagnosis inference method is proposed based on the integration of qualitative reasoning and quantitative simulation. Simulation results with the test-fire data of complicate systems show that, the method proposed can eliminate ambiguity in reasoning results effectively and thus improves the performance of qualitative reasoning methods.
出处
《系统仿真学报》
CAS
CSCD
2003年第5期689-692,共4页
Journal of System Simulation
基金
国家自然科学基金资助(NR.59806014)
关键词
定性推理
定量仿真
故障诊断
知识
qualitative reasoning
quantitative simulation
fault diagnosis
knowledge
