Intelligent Wireless Fire Monitoring System Design And Indoor Fire Source Localization Method Research

In recent years, with the constantly increasing the height of the floor of themodern urban architecture, the building density is increasing, making the indoor fireincidents becoming more frequent. And the property damage and personal injury werecaused by more and more serious. As a result, we need much higher demands for themodern building fire system technology. Thus, carrying out the research of intelligentwireless fire-fighting monitoring system design and indoor fire source locationmethod has important theoretical and practical values, in particular, that wouldproduce good social effects.There are two key issues in technology research and development of intelligentfire-fight systems. One is that the building fire-fighting detection system to thedirection of intelligent wireless，another is how to better achieve the fire sourcelocalization in early stage of a fire incident. Therefore, this paper carried out from twoaspects of building the smart wireless fire monitoring system and indoor fire sourcelocalization method. This paper main study the design of intelligent wireless firemonitoring system and two algorithms of indoor fire source localization; specificresearch work and contributions are as follows:1) Design an intelligent fire monitoring system base on wireless sensor network.Base on the latest development of the wireless sensor network technology and firedetection alarm technology, we use composite sensor as fire detection node and use ofwireless sensor networks to build the intelligent wireless fire monitoring system. Atlast the whole system design is given, and giving a simple experiment test.2) Establish the fire source clustering positioning method based on dynamicclustering idea with the sensor arrays. The output of the traditional fire sourcelocalization method based on wireless sensor array was randomness and uncertainty,which cannot be a better guide to actual fire-fighting. First this paper gives a denseregion selection algorithm based on dynamic clustering idea, and then selects thedense region of each output of the traditional methods, and finally each selected areais estimated as a fire source region and the regional centers as the point estimate of thefire location.3) Proposed a fire source localization method base on statistics with the sensorarrays. First, to get the flue gas propagation delays by using Grey Relational Analysisto the collected temperature signal within the sensor array; Then, get the azimuthestimates according to the propagation model and the geometric relationship;Furthermore, based on statistical modeling approach azimuth estimation as a meanand variance to characterize the random variable, which estimated the fire sourceconversion is estimated that the regional context of the fire source; Finally, two sets ofcross-array way to build estimated fire source location regional, and the regionalcentre as a point estimate of the fire location.