The Application Of Fuzzy Neural Network In Fire Signal Recognition

The development of fire signal detection system is analyzed in this paper. the emphasis is put on the study of the complementary relation between fuzzy logic and artificial neural network. We find that an ANN can be initialized by the experience of a fuzzy system and a fuzzy system can be constructed by an ANN. by analyzing the shortage of the traditional fire signal detection method and the specialty of fuzzy logic and artificial neural network, we establish a fire signal detection model based on fuzzy neural network, and present its learning algorithm. It is demonstrated by computer simulation that the algorithm has many merits, such as: learning quickly, converging asymptotically. based on this model, a fire signal detection method is put forward. Digital filter algorithm of this system is designed. It insures that the fire forecast is real-time, exact and reliable. And it increases the anti-jamming ability of the systemand the interface of MCU is studied. Using today's electronic technology, we design a hardware circuit based on MCU to actualize the method. It is demonstrated by experiment that the method has strong ability of anti-jamming and low rate of misinformationfurthermore, the model is very multipurpose. It can be used in other fields, such as error detection, by adjusting its parameters.