Realization of Fault Diagnosis Rule Extraction System Based on Rough Set Theory

摘要

This paper introduces the implementation process of the fault diagnosis rule extraction system based on rough set theory. In order to extract the rules, the system imports the original fault diagnosis decision table firstly, and then reduces the attribute set by distinct function and distinct matrix. Based on the reduced matrix, system constructs a various simplified level decision network, and then takes the reduced attribute set as the start node of this network. It gets the decision rule set according to the decision network. At the same time, to filter the noise effectively, the system introduces the coverage degree concept in the rule extraction system on the basis of the accuracy degree and filters the extracted rules. At last, it extracts the effective diagnosis rules. A hydraulic control system fault diagnosis rules extraction demonstrates the system operation process.