Application rough sets to safety risk index screening in ATC

Index system construction of ATC safety risk usually based on hierarchy subdivision, it is easy to exist correlation and redundancy between various levels indicators. In order to reduce the index drift values caused by subjective factors, Using rough sets which are advantage in dealing with the aspects of fuzzy and uncertainty, through the methods of attribute reduction algorithm to eliminate redundant index, it can realize warning monitoring to the key risk object which affects system state. After an application example screening from primary index, it effectively reduced redundant index, and the key index comply with the system state characteristic.