基于多信息融合的电力火灾综合探测技术研究

发布时间：2018-03-03 21:40

本文选题：电力　切入点：火灾探测　出处：《消防科学与技术》2017年02期 　论文类型：期刊论文

【摘要】：设计了包含不同火灾特征参数传感器的吸气式火灾复合探测器。以多传感器动态过程信息作为火灾判断的数据基础,采用BP神经网络方法对多路探测信号进行融合处理。开发了电力火灾在线探测系统,并应用WebSocket技术,将火灾探测及报警信息快速推送给用户。实验结果表明:电力火灾综合探测系统无误报漏报情况,大大提高了火灾判断的准确率。火灾报警时间显著提前,而采用服务器推送技术可将报警时间进一步缩短,提高了火灾响应速度。
[Abstract]:An air-breathing composite fire detector with different fire characteristic parameter sensors is designed. The multi-sensor dynamic process information is used as the data basis of fire judgment. The method of BP neural network is used to fuse the multi-channel detection signal. The on-line detection system of electric power fire is developed, and the WebSocket technology is applied. The information of fire detection and alarm is quickly pushed to the user. The experimental results show that the integrated detection system of electric fire reports the missing information correctly, which greatly improves the accuracy of fire judgment, and the fire alarm time is significantly advanced. The server push technology can further shorten the alarm time and improve the fire response speed.
【作者单位】：国网安徽省电力公司电力科学研究院;国网安徽省电力公司;
【基金】：安徽省自然科学基金项目(1408035MKL94);2016中国消防协会科学技术年会“青年消防学者论坛”交流论文
【分类号】：X932

【相似文献】