摘要
为实现对火灾现场安全的精准监控,文章提出基于多传感器融合的建筑物火灾安全监控方法。该方法引进温度传感器、CO传感器、烟雾传感器,采用不确定推理(Dempster-Shafer, D-S)证据理论,对多传感器采集的多源信息进行融合,建立建筑火灾发生信任函数;利用反向传播神经网络(Back Propagation Neural Network, BPNN),将传感终端获取的融合信息输入BP神经网络,进行火灾信息的修正;将速率检测算法引入局部决策,通过将测量到的信号变化率与预先设定的固定阈值进行对比,从而实现对火灾的判定与安全监控。结果表明,该设计方法能够对建筑物火灾进行精准预警,还能实现火灾现场的温度、烟雾浓度的精准监测,保证监控结果的准确性与可靠性。
To achieve precise monitoring of fire scene safety,this study proposes a building fire safety monitoring method based on multi-sensor fusion.Introducing temperature sensors,CO sensors,and smoke sensors,using the Dempster-Shafer(D-S)evidence theory to fuse multi-source information collected by multiple sensors,and establishing a trust function for building fire occurrence;Using the Back Propagation Neural Network(BPNN)neural network,the fusion information obtained by the sensing terminal is input into the BP neural network for the correction of fire information;Introducing rate detection algorithms into local decision-making,by comparing the measured signal change rate with a pre-set fixed threshold,fire detection and safety monitoring can be achieved.The results indicate that the design method can provide accurate early warning for building fires and achieve precise monitoring of temperature and smoke concentration at the fire scene,ensuring the accuracy and reliability of monitoring results.
作者
徐刚
鲁英奇
张佰丰
XU Gang;LU Yingqi;ZHANG Baifeng(Shanghai Environmental Protection(Group)Co.,Ltd.,Shanghai 200433,China;Shanghai Anting Wastewater Treatment Co.,Ltd.,Shanghai 201814,China)
出处
《无线互联科技》
2025年第20期88-91,共4页
Wireless Internet Science and Technology
关键词
多传感器融合
火灾信息修正
信任函数
安全监控
火灾
建筑物
multi-sensor fusion
fire information refinement
trust function
safety monitoring
fire
building
