| The fire alarm automatic system could find fire timely, start other fire-fightingequipments, and control fire detecting, alarming and extinguishing automaticly. It is vital tothe modern buildings for fire safety. As technology of the sensor, the wireless communication,the artificial intelligence and computer test/control gradually progresses, the automatic firealarm system meets an excellent opportunity. The purpose of this study was to improve theintelligence level, and reduce the wrong alarm rate and failure alarm rate of the automatic firealarm system.Firstly, this paper analyzed the wireless monitor network and its application at home andabroad, and selected ZigBee wireless monitor networks suitable for fire alarm automaticsystem, then a wireless automatic fire alarm system communications network was designedand set up in3floor of building17in Tianjin University of Technology. Secondly, analysis ofthe fire model characteristic was described. The BP neural network intelligent algorithm withits high nonlinear mapping, self-learning, parameter parallel processing, fault tolerance abilityhad been applied to fire detection. This algorithm deals with the multiple fire parametersfirstly, and then judges the fire situation. Thirdly, the background of Virtual Instrument andthe technical characteristics of LabVIEW were introduced. BP neural network algorithm and aset of intelligent fire alarm automatic system had been set up by LabVIEW Virtual Instrument.Finally, two fire identification models were designed, and carried on simulation and fire test.Test proved that BP neural network algorithm can validly settle the contradiction betweensensitivity and accuracy, and obtain anticipative purpose basically.This wireless intelligent fire alarm automatic system has the features of high reliabilityand intelligence, but it has some part that need be improved. So paper finally puts forwardsome methods, and prospects the future of wireless intelligent fire alarm system. |
PDF Full Text Request