A Decision-Making Method for Fire Detection Data Fusion Based on Rough Set Approach

This paper introduces an attribute reduction method based on the characteristics of consistent approximation space (CAS) of rough set. The proposed method can be used to fuse alarm signals from different kinds of fire sensors using decision rules derived from a sample data in CAS with the data classification features of CAS. The experiment results of an example data set show that this method can fuse different kinds of fire alarms to detect the occurrence of fire as well as the type of fires and give the confidence of the decisions.