基于BP神经网络的多信号融合智能消防报警模型研究  

Multi Signal Fusion Intelligent Fire Alarm Model Based on BP Neural Network

作　　者：朱晟文 ZHU Shengwen(Hefei Fire and Rescue Bureau,Hefei,Anhui Province,230000)

机构地区：[1]安徽省合肥市消防救援局,安徽合肥230000

出　　处：《长江信息通信》2025年第2期137-139,共3页Changjiang Information & Communications

摘　　要：针对传统智能消防产品功能与通信方式单一与运行功耗较高的问题,研究设计一种联合物联网技术、故障诊断技术、危险气体浓度检测、温度检测与红外火焰检测等多种功能的多信号融合智能消防报警模型,并在此基础上,通过误差反向传播神经网络处理数据与分析数据,并与智能模型中的单片机产生连接。研究结果表明,研究方法在实际火灾应用场景的综合火焰检测率为97.5%,并且温度采集精度高达99.62%。上述结果说明研究方法具有更多样化的功能与通信方式,并且性能得到显著增长,促进了智能消防的发展。A multi signal fusion intelligent fire alarm model that combines multiple functions such as Internet of Things technology,fault diagnosis technology,hazardous gas concentration detection,temperature detection,and infrared flame detection is studied and designed to address the problems of single functionality and communication methods,as well as high operating power consumption in traditional intelligent fire products.Based on this model,data is processed and analyzed using an error backpropagation neural network,and connected to the microcontroller in the intelligent model.The research results indicate that the comprehensive flame detection rate of the research method in practical fire application scenarios is 97.5%,and the temperature acquisition accuracy is as high as 99.62%.The above results indicate that the research method has more diverse functions and communication methods,and its performance has significantly increased,promoting the development of intelligent fire protection.

关 键 词：BP神经网络 多信号融合 消防报警 智能模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]