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。
Aspirating compound fire detector including different fire feature parameters＇ sensors was designed. Multi-detecting- signals based on the fire judgment of the dynamic infoi＇mation from multi-sensors were fused through BP neutral network. The online detecting system of electric power fire was devel- oped, and the information of fire detecting and alarm were quickly pushed to users through Web Socket technology. Ex- periment indicated the integrated detection system of electric power fire has no false positives and false negatives, which greatly raises the accuracy of fire judgment. The fire alarm time is advanced obviously, while it is advanced further after adop- ting server push technology, which raises the speed of fire re- spond.
出处
《消防科学与技术》
CAS
北大核心
2017年第2期213-217,共5页
Fire Science and Technology
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
关键词
电力
火灾探测
多传感器
BP神经网络
electric power
fire detection
multi-sensors
BP neutral network
