无线多传感器信息融合的火灾检测系统设计  

Design of Fire Detection System Based on Wireless Multi-sensor Information Fusion

作　　者：曾思通[1] 童晓薇[1] 沈培辉[1] ZENG Sitong;TONG Xiaowei;SHEN Peihui(Department of Mechanical Engineering,Fujian Chuanzheng Communications College,Fuzhou Fujian 350007)

机构地区：[1]福建船政交通职业学院机械工程系

出　　处：《湖北理工学院学报》2019年第6期23-27,32,共6页Journal of Hubei Polytechnic University

基　　金：福建省自然科学基金项目(项目编号:2015J01186);福建省教育厅科技项目(项目编号:JZ180346;JZ180373)

摘　　要：为了精确高效地检测识别火灾,克服传统火灾检测器检测信息单一、布线复杂、环境适应能力差等缺点,设计了一种基于ZigBee无线通信的多传感器信息融合的建筑火灾检测系统。该系统采用烟雾传感器、温度传感器和CO传感器作为火灾检测传感器终端,通过ZigBee无线模块将火灾数据传送至主控器STM32F407单片机进行离线检测,在MATLAB上利用模糊神经网络对传感器数据进行融合决策,最后将融合算法移植至STM32F407单片机进行离线检测。选用常见火灾材料及干扰源进行真假火源试验,结果表明,开发的系统具有较高的检测效率和检测精度,可以实现建筑火灾的检测识别。A building fire detection system based on multi-sensor information fusion of ZigBee wireless communication was designed to detect fire accurately and efficiently, which can overcome such shortcomings of traditional fire detection system as single detection information, complicated wiring and poor environmental adaptability. Smoke sensor, temperature sensor and CO sensor were used to collect the field data of the fire monitoring area. Then the fire data was transmitted to the main controller STM32 F407 through the ZigBee wireless module. The fuzzy neural network was used to fuse the sensor data on MATLAB, and the fusion algorithm was transplanted to STM32 F407 finally. Then the common fire materials and the interference sources were used to carry out true and false fire resource tests. The experimental results showed that the developed system had highly detection efficiency and detection accuracy, which can realize the detection and identification of building fires.

关 键 词：火灾检测 ZIGBEE 数据融合 神经网络 STM32F407 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]