The Study On Fire Alert Detecting System In Olfactory Analogy

With the development of human society fire is very important to people. However if the fire is out of control, it will bring human a disaster. We call "fire disaster" . Fire not only deprives human's lives and estates, but also destroys environment and nature resources which support all people in the world. Therefore, it is a forever task for human to prevent fire breaking out and reduce losing owning to fire.So the study of fire detecting technology is more important. According to detecting of certain single fire detector, the comparative law of threshold is adopted to alarm fire in traditional fire detecting technology. But it is very usual to mistake-alarm due to the contradiction between diverse fire signals and single detector type. In order to reduce mistake-alarm, at present, many researchers are devoted to using intelligence-detector instead of single sensor to distinguish non-fire signals and genuine fire signals.The olfactory techniques based on gas sensor array and pattern recognition and pre-processing standardization of gas sensor array signals. Scientists apply olfactory techniques to fire detecting system, they finds that there are many characteristics: first, this technique can process diverse fire signals and give a satisfaction result, it can improve reliability; second, these techniques include many advanced algorithms such as fuzzy-neutral algorithm. It can utilize the rule of fire distinguishes; final, Multiple messages extraction technology can transfer non-linear to linear.If only "and-not" judge processes multiple-sensor signals, it's equal to be a combination of plus and minus. That can't make full use of the function of multiple-sensor. The gas sensor array in olfactory techniques can make full use of the function of multiple-sensor. It utilizes various kinds of sensors information to draw the overall, correct understanding of the environment or target's characteristics. The theoretic in this thesis consists of three parts: how to determine parameters of fire signals; the fire detecting principles and algorithms; design the fire detecting motherboard electric circuit design.This paper analyze the characters of fire and olfactory technology, the author proposes a kind of algorithm using data fusion technology in view of artificial intelligence. The input analog signals are provided by temperature, smoke, carbon monoxide sensors, and the algorithm synthesizes those signals by use of neural network and fuzzy logic technology. The aim is to design a kind of quick, exact and effective fire detecting system, which can reach the aim of early alarm and reducing mistake-alarm.