摘要
为进一步提高火灾探测系统的识别精度,设计了温度-火焰复合探测系统。依据特种火灾探测器国家标准,采集了大量试验数据,分析了目前广泛应用的火灾探测算法,建立了串联型模糊神经网络,实现了基于模糊神经网络的火灾识别。仿真结果表明,与单独采用BP神经网络相比,采用模糊神经网络将识别率提高了3%,该方法可有效应用于火灾探测中。
In order to further enhance the identification accuracy of the fire detection system,the compound detection system based on temperature and flame has been designed.According to the National Standard for Special Fire Detectors,large amount of test data are acquired.The fire detection algorithms commonly used at present are analyzed;and serial fuzzy neural network is established;thus the fire identification based on fuzzy neural network is implemented.The results of simulation indicate that comparing with using BP neural network,the recognition rate can be increased 3% by the method proposed.The method can be effectively used in fire detection.
出处
《自动化仪表》
CAS
北大核心
2011年第10期48-51,共4页
Process Automation Instrumentation
基金
国家自然科学基金资助项目(编号:60975079)
上海市教育委员会创新基金资助项目(编号:11YZ19)
关键词
火灾探测
模糊神经网络
人工智能
自动控制
信号处理
标准
Fire detection Fuzzy neural network Artificial intelligence Automation control Signal processing Standard
