首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种基于前馈神经网络的火灾探测方法
引用本文:杨宗凯,王殊,何建华,沈春蕾.一种基于前馈神经网络的火灾探测方法[J].华中科技大学学报(自然科学版),1997(2).
作者姓名:杨宗凯  王殊  何建华  沈春蕾
作者单位:Yang Zongkai Assoc. Prof.; Dept. of Electronics & Information Eng.,HUST,Wuhan 430074,China. Wang Shu He Jianhua Shen Chunlei
摘    要:讨论了基于模拟量探测器和前馈神经网络的火灾探测方法.以光学感烟、半导体感温、离子感烟和湿度模拟传感量为输入,以火灾概率和阴燃火灾概率为输出量对前馈神经网络进行训练,不仅可提取火灾环境的有效特征,而且可提高检测算法的可靠性

关 键 词:火灾报警  神经网络  信号处理

A Fire Detecting Method Based on the Feedforward Neural Network
Yang Zongkai Assoc. Prof., Dept. of Electronics & Information Eng.,HUST,Wuhan ,China. Wang Shu He Jianhua Shen Chunlei.A Fire Detecting Method Based on the Feedforward Neural Network[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,1997(2).
Authors:Yang Zongkai Assoc Prof  Dept of Electronics & Information Eng  HUST  Wuhan  China Wang Shu He Jianhua Shen Chunlei
Institution:Yang Zongkai Assoc. Prof., Dept. of Electronics & Information Eng.,HUST,Wuhan 430074,China. Wang Shu He Jianhua Shen Chunlei
Abstract:A fire detecting method is proposed based on an analog detector and feedforward neural network. The feedforward neural network is trained with the optic smoke sensing, semiconductor temperature sensing, ion smoke sensing and humidity analog as input and the fire probability and smouldering fire probability as output. Not only can the effective features of the fire environment be extracted but also the reliability of the detecting algorithm is improved. The results will be applied to the first generation of the intelligent automatic fire detecting and warning system with neural network.
Keywords:fire warning  neural network  signal processing  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号