Intelligent Fire Auto-detection Technology Base On Fuzzy And Neural Network

Combining with the character of fire alarm system, this paper puts forward a fire detection model and an intelligent arithmetic that can present a fire prediction probability on the base of summarizing neural network and fuzzy logic. The arithmetic combines with the character of neural network and fuzzy logic. It judges on the detection signal changing trend in a period. So, the arithmetic could present fire prediction probability precisely, and simulation research proves the method feasibility. The paper designs and debugs a hardware interface board on the base of the fire intelligent alarm arithmetic. The board fulfills the exchanging data function between industrial controlling computer and outside. The board bus standard is ISA, communicating according to 485 protocol. The paper fulfills the board locking work mode. The paper programs the software of fire alarm system. Through interface board, the software could get each ioop detector information and response information to software interface. At the same time, through interface board, user could send out various commands to control any detector running status. The paper fulfills part key module programs on the base of software whole frame, i. e. communication module, database operation module and so on. Experiments prove the rationality of design.