Design Of Flame Detection System Based On Multi-infrared Bands Sensor

With incredible quick development of our modern society, it becomes more and more urgent to find a quick and efficient way to keep the possible fire hazards under surveillance. The earliest detection of possible fire hazards surely prevents fire hazard or greatly reduces the loss of property and casualties. Obviously, it's very important for the skyscrapers, big warehouses, performance and entertainment centers and crowded public gathering locations to have fire detection system to prevent fire hazards. Therefore, it is very essential to develop an accurate and quick-respond fire detection system.This thesis focuses on how to devise a fire detection system by using PerkinElmer's component TPS2534, a dual thermopile sensor and TI's component MSP430F149, a microprocessor unit, which quickly spots the heat discharged by infrared light and immediately gives out warnings. This flame detection system responds more quickly to visible flame. Compared with traditional heat-sensitive and smoke-sensitive fire detection system, its warning accuracy is higher.This paper mainly deals with how TPS2534, a dual-thermopile sensor works in fire detection system, and basic ways in which flame detection system works. And advantages of this fire detection system are presented by comparison with traditional ones. In this paper, hardware design is also elaborated. Mainly, how MSP430, the algorithm controlling core and TPS2543, the dual-thermopile sensor work and systematic design are elaborated. The peripheral modules, such as memory module, real time module are presented. In addition, recognition algorithm is clarified to make it more comprehendible. Cross correlation algorithm applied in this fire detection system is presented here, and the advantages and disadvantages of the algorithm are analyzed compared with other algorithms.