摘要
研究准确识别飞机火警系统,由于飞机火灾火势猛,突发性强,飞机火警系统误报率较高。为提高识别火警准确率,提出一种D-S证据理论的数据融合方案。对机舱内燃烧时的温度、烟雾浓度、光亮度等火情特征参量进行判断,分析特征参量的内在特性和常见的火情信息处理流程。采用多传感器多周期的时空融合方法对火情信息进行识别,得出火灾的四种不同状态类型:明火、阴燃火、无火灾、无法识别,并进行仿真,结果表明,根据D-S理论进行数据融合结果对火警识别具有较高的准确度和可信性,为设计提供了依据。
For high rate of false fire alarming in the airplane, data fusion was proposed based on D-S theory. Temperature, smoke density, lightness and other fire features were described, which were produced by some material burning in the airplane. Some inner features were analyzed, and common fire information processing flow was given. Information from multi-sensors was fused with different measurement cycles, which got four types of results： open fire, smoldering fire, no fire and no identification. The result of simulated experiment shows that the application of temporal-spatial information fusion in fire alarming brings higher accuracy and reliability.
出处
《计算机仿真》
CSCD
北大核心
2012年第1期71-74,86,共5页
Computer Simulation
基金
江苏省高校自然科学基础研究项目(08KJD520011)
关键词
火灾探测
证据理论
数据融合
Fire detection
Evidence theory
Data fusion
