机电系统故障诊断知识描述与生成技术的探索

Presentation and Generation of Fault Diagnosis Knowledge for Mechano-Electronic Systems

目的研究机电系统概念描述和故障诊断知识的表达、获取及生成技术。方法提出基于概念的、应用结构化语言描述物理系统的方法。将概念分为实体概念、属性概念和特征概念，在实体概念的基础上构造概念词典和概念树．通过概念树描述机电系统内部结构．综合上述３种概念描述机电系统的故障信息以及故障诊断知识。结只针对ＦＭＳ系统的故障诊断，建立了从概念描述、知识表达到知识生成全过程的系统框架，建立了包含维修数据库、知识库和推理机一体化的故障诊断和维修系统。结论该方法是可行和有效的，所建故障诊断系统具有一定的学习能力。

Aim To focus on the concept description and fault diagnosis knowledge representation, acquisition and generation for complex mechano-electronic-electronic systems. Methods A concept-based structured language was proposed to describe the practical system. The concepts were classifted into three categories as entity concept, attribute con- cept and feature concept. Based on the entity concept, a concept dictionary and some concept trees were created. The inner structure of a complex mechano-electronic system was expressed through the concept tree. The fault information and fault diagnosis knowledge for the system were represented by the combining the three kinds of con- cepts mentioned above. Results Based on which a framework for FMS troubleshooting has been established. The framework encompasses the entire process from concept description, knowledge representation, to knowledge generation. An FMS fault diagnosis and maintenance system, in which a maintenance-database, an image-base are integrated with knowledge-bases, is established. Conclusion The meth- od is feasible and effective,and the fault diagnosis system is able to learn.