飞机货舱复合烟雾探测方法研究

Research on multi-sensor smoke detection method for aircraft cargo compartment

为解决目前飞机货舱火警探测装置易受空气中灰尘和水蒸气颗粒干扰造成误报率极高的问题,提出一种复合多传感器的飞机货舱火灾探测方法。首先搭建包括温度传感器、CO传感器和双波长光电式烟雾传感器的复合式火灾探测装置,设计火灾探测系统软件;然后进行真假火源试验,采集火灾过程中烟雾、温度、气体的变化特征参数;最后采用人工神经网络算法对采集到的数据进行融合分析。结果表明:嵌入双波长光电式烟雾探测器的多传感器探测装置报警正确率比传统火灾烟雾探测器有大幅度提高,干扰源识别相对误差不超过5.7%。

In order to solve the problem of the high false alarm rates of the aircraft cargo compartment fire detectors caused by the dust and water vapor particles in the air,a multi-sensor composite aircraft cargo compartment fire detection method was worked out.Firstly,a composite fire detection device was built,including a temperature sensor,a CO sensor and a dual-wavelength photoelectric smoke sensor.A fire detection system software was designed.Then a large number of true and false fire source experiments were carried out to collect data on the parameter variation features of smoke,temperature and gas during the fire.Finally,the artificial neural network algorithm was used to perform fusion analysis of the collected data.The experimental data show that the alarm accuracy of the multi-sensor detection device embedded with dual-wavelength photoelectric smoke detector is significantly higher than that of traditional fire smoke detectors,and the relative error of interference source identification does not exceed 5.7%.