同类型易燃物集中区域火情预警报警的关键技术研究

发布时间：2018-06-08 00:14

本文选题：火情预警 + 火情定位　；参考：《南京信息工程大学》2017年硕士论文

【摘要】：火灾给人类社会带来的灾难数不胜数,特别是同类型易燃物集中区域火灾的火势相当猛烈,不易扑救。本文探索出一套成熟的解决方案,能够在第一时间发现火源,并及时预警报警。本文所做的工作主要是如下几个部分:(1)本文根据多位著名科学家的研究成果,借鉴红外制导空空导弹的成熟经验,提出了基于碲镉汞红外传感器并可用于各种同类型易燃物集中区域的火情感知方案。本文还将非常成熟的无线电测向思想进行跨界转化,利用多台火情探测设备上报的火情点相对探测设备方向角数据进行火情定位的方案,解决了当下解决方案无法对室外场所进行有效火情定位的弊病。(2)本文为同类型易燃物集中放置的大跨度仓库设计了更为成熟的多传感器数据融合分析解决方案,该方案充分利用BP神经网络和ART神经网络自身的优势,结合前期提取的特征信息,严格参照国家相关标准,实现了精度和效率的双提升。同时,利用一个CPU和多个GPU在一台普通工作站上针对BP神经网络速度慢的缺陷进行并行并行计算改造,在提升计算效率的同时降低了硬件成本。(3)鉴于没有现成的火情探测设备支撑本文提出的多项独特的关键技术设计,本文根据创新设计方案给出了火情探测设备完整的设计方案并对关键组成部分的电路进行了详细设计和说明,最终得到了完整的软硬件成果。
[Abstract]:Fire brings numerous disasters to human society, especially the fire in the same type of combustible concentration area is very fierce, not easy to fight out. This paper explores a set of mature solutions, can find the source of fire in the first time, and early warning alarm. The work done in this paper is mainly as follows: 1) based on the research results of many famous scientists, this paper draws lessons from the mature experience of infrared guided air-to-air missiles. A fire sensing scheme based on HgCdTe infrared sensor is proposed, which can be used in various concentrated areas of the same type of flammable substances. In this paper, the very mature idea of radio direction finding is also transformed across the border, and the fire location scheme is implemented by using the fire point relative to the direction angle data of the detection equipment reported by several fire detection equipment. This paper designs a more mature multi-sensor data fusion analysis solution for the large span warehouse with the same type of combustible materials. This scheme makes full use of the advantages of BP neural network and art neural network, combines the feature information extracted in the early stage, strictly refers to the relevant national standards, and realizes the double promotion of accuracy and efficiency. At the same time, using one CPU and several GPUs in a common workstation to solve the problem of the slow speed of BP neural network, the parallel computation is carried out. In view of the fact that there is no ready fire detection equipment to support the unique key technology design proposed in this paper, In this paper, according to the innovative design scheme, the complete design scheme of fire detection equipment is given, and the circuit of the key components is designed and explained in detail. Finally, the complete software and hardware results are obtained.
【学位授予单位】：南京信息工程大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP277;TU892

【参考文献】