Extracting optimal generalized decision rules for fault diagnosis from incomplete data based on rough set

In fault diagnosis, the extraction of rules from incomplete data is usually more difficult than from complete data. By means of the new definition of discern ability matrix primitive, a method for directly extracting optimal generalized decision rules for fault diagnosis from incomplete data based on the rough set is proposed. By using the maximal consistent blocks as column units to construct the discern ability matrices, all object-oriented reductions are found and all optimal generalized decision rules for fault diagnosis from incomplete data are extracted. The method proposed does not require a change in the size of the original incomplete data set, and has higher efficiency of computing reduction. The application of the method is demonstrated with a fault diagnosis example of the operational state of an electric system with incomplete data. The validity of this method is proved.