Knowledge Discovery for Oil Diagnosis System Based on Rough Set
Jin Wang · Tribology · 2003
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 fourinput system to a twoinput one.