基于颗粒光散射矩阵特性的民用飞机货舱低误报率烟雾探测研究

Currently, photoelectric smoke detectors are commonly used for fire alarm in civil aircraft cargo compartments. However, the high false-alarm-rate of this type of detector affects the safe operation of civil aircraft seriously..The project aims to improve the performance of photoelectric smoke detectors based on scattering matrix analysis with study of the light scattering mechanism of particles. Smoke particles and nuisance aerosol particles will be sampled based on experiments in a full size aircraft cargo compartment simulator. The morphology, size distribution, chemical elemental composition and other characteristics of sample particles will be analyzed. The scattering models of typical fire smoke particles and nuisance particles will be established based on their microstructure and microscopic physicochemical properties. The scattering matrices of particles will be calculated by discrete dipole approximation (DDA) and other numerical simulation methods. The scattering matrices of sample particles will be measured with the platform of a polarization-modulated multichannel Mueller-matrix scatterometer for particles. A comparative analysis between experimental and calculated results will be carried out to verify the scattering models of typical fire smoke particles and nuisance particles in aircraft cargo compartments. The optimization method of key parameters, such as scattering matrix elements, wavelength, and the scattering angle, will be studied based on the experimental and simulation results. A recognition algorithm which can respond uniformly to all kinds of fire smoke particles and with which nuisance particles can be distinguished will also be developed. The prototype devices will be designed and verified in the aircraft cargo compartment simulator.

目前民用飞机货舱火警探测普遍使用的光电感烟探测器存在误报率高的缺陷，严重影响着民用飞机的安全运营。. 本项目试图从颗粒光散射机理出发，利用光散射矩阵分析的方法实现光电感烟探测器的性能改进。通过全尺寸模拟实验采集飞机货舱典型火灾烟雾颗粒和干扰颗粒，测量分析其微观形貌、粒径分布以及化学元素组分等特征。基于颗粒微观结构和理化性质等特征信息建立颗粒的光散射模型，采用离散偶极子近似等方法对颗粒的光散射矩阵进行模拟计算。利用光散射实验平台测量颗粒的光散射矩阵，并与计算结果进行对比分析，验证货舱环境下典型火灾烟雾与干扰颗粒群的光散射模型。基于实验和模拟结果，研究散射矩阵元素、光源波长、散射角度等关键参数的优化方法，发展对火灾烟雾颗粒响应一致并区分干扰颗粒的识别算法，研发样机装置并在货舱模拟环境内进行验证。

针对目前民用飞机货舱火警探测普遍使用的光电感烟探测器存在高误报率的问题，本项目采用光散射矩阵分析的方法为光电感烟探测器性能的提升提供基础研究支撑。.项目研究了不同压力条件下典型货舱可燃物燃烧特性与燃烧产物特征，获取了粉尘等典型干扰颗粒的化学组成、形貌与粒径分布特征。实验测量了不同波长下典型火灾烟雾颗粒与干扰颗粒光散射矩阵元素的角度分布，揭示了火灾烟雾颗粒与干扰颗粒光散射矩阵元素随角度的变化规律及差异性特征。基于颗粒物的形貌、粒径分布、折射率等参数，分别为烟雾颗粒与干扰颗粒建立了颗粒光散射模型，对单颗粒和颗粒群的光散射矩阵进行了计算，建立了典型火灾烟雾颗粒与干扰颗粒光散射矩阵数据库。提出了单一散射角度的单一元素比、不同散射角度的单一矩阵元素比、同一散射角度的不同矩阵元素比等三类可用于火灾烟雾颗粒与干扰颗粒识别的判断依据，并开展了基于偏振光散射的，具备颗粒识别能力的光电感烟探测器原理样机设计。项目共计在国内外重要期刊和会议上发表论文20篇，申请专利5项，完成博士学位论文2篇，硕士学位论文2篇，项目负责人参与了民航局重点实验室“民航热灾害防控与应急重点实验室”的申报并成功获批。项目成果为我国和国际上民用飞机货舱光电感烟探测器防误报能力的改进提供了重要的基础数据和方法支撑。