基于Rough Set的油液故障诊断系统的知识发现

Knowledge Discovery for Oil Diagnosis System Based on Rough Set

结合RoughSet理论和摩擦学系统的特点,讨论了油液故障诊断系统的不协调性.在包含度方法的基础上,将普通二元关系进行推广,提出了一种不协调油液故障诊断系统知识发现模型,给出具体的运算方法,并通过试验实例验证了该模型的有效性.结果表明,该模型在最大分布约简的基础上进行油液诊断知识获取,能够很好地完成不确定性问题的推理,并且可以推导出具有最大可信度的油液诊断知识规则.

The inconsistency of knowledge discovery for oil fault diagnosis system (OFDS) was discussed based on Rough Set theory and tribological system features. A knowledge discovery model for inconsistent OFDS was proposed making use of inclusion degree method and expanding general binary relation. It was suggested to use maximum distribution reduction method for knowledge discovery in the operation algorithm, while an application example was given to demonstrate the validity of the established model. As the results, it was feasible to carry out uncertain reasoning and convenient to acquire oil fault diagnosis knowledge rules with maximal reliability making use of the model. Moreover, it was able to carry out the operation of the spectrometric data of a diesel engine oil with greatly simplified knowledge using the model which made it possible to transform a fourinput system to a twoinput one.