Method of fuzzy neural network for fire detection

A fuzzy neural network for fire detection is presented according to the cha racteristics of fire signals, and a calculating model of fuzzy inference is proposed to process fire signals. Fuzzy operations are applied to analyze synthetically different kinds of fire signals. Then, a calculating model of learning algorithm is created for a set of membership functions and fuzzy rules. The fuzzy neural network, which combines the techniques of neural ne twork and fuzzy logic, offers a signal processing method for fire detection with fuzzy and adaptive characteristics simultaneously.