CO和CO2气体红外光谱技术在火灾早期探测中的应用研究

通过对现有各种火灾探测方法进行归纳分析,提出以CO和CO2作为探测特征参量的傅里叶红外-光谱技术早期火灾探测新方法,并在此基础上建立了完善的早期火灾实验系统.通过大量真假火灾实验得出常见材料在火灾发生过程中产生气体浓度的变化特征:真实火灾产生出大量的CO和CO2,浓度变化非常有规律;假火灾只产生极少量的CO,且浓度变化杂乱无章.从浓度数据中提取出火灾发生的特征信息制定探测算法就可以有效地提高早期报警的分辨率,降低火灾探测的误报率.     

基于图像探测的早期火灾烟气模式研究

烟气是火灾过程中的重要物理现象。为改进早期火灾探测和预警,需要对基于图像探测的火灾烟气模式加以研究。以不同应用背景为依据,对烟气模式提出分类研究的思想。并以提高图像探测的可靠性和灵敏度为目的,对烟气的图像信息特征进行分析,提出了建立早期火灾烟气模式的设想。     

基于多传感器信息融合的火灾报警器设计

火灾是人类社会经常遇到的一个灾害,它通常给人民群众带来生命和财产的损失。针对传统火灾报警器参数单一,易出现漏报、误报的问题,本文设计了一种基于多传感器信息融合的火灾报警器;该系统以温度传感器、CO浓度传感器和烟雾浓度传感器测量的现场参数为参考,通过比较器的比较送给单片机进行信息融合,当温度高度设定值时,蜂鸣器报警;在温度高于设定值,且CO和烟雾浓度高于设定值时,蜂鸣器报警并且LED灯闪烁,提示有可能有火灾发生。最终通过对系统的调试实现了对液晶显示器的显示控制和报警器的报警。仿真结果表明了系统能够及时准确的判断是否有火灾发生,具有较强的火灾预测和报警能力。     

目标表面发射率对红外辐射偏振特性的影响分析

偏振探测不仅可以获得目标的光谱、强度、偏振态以及空间几何形状等参数,还可以获得更丰富的目标信息,有利于改善对目标的探测和识别能力。红外辐射偏振成像是近年来发展起来的一种新的红外探测技术,主要通过目标与场景的红外辐射偏振特性差异进行目标探测与识别。但由于红外辐射偏振信息在传输过程中,其偏振态会受到传输介质的影响,而通常基于实验分析总结目标红外偏振特性的方法难以对偏振传输过程中的影响因素进行估计,也不能定量描述各相关参数对于红外偏振信息的影响。通过微面元理论的双向反射分布函数模型,建立了基于偏振双向反射分布函数的红外辐射偏振传输方程, 推导分析了目标表面发射率对红外辐射偏振度的影响,结果表明：目标表面发射率对目标红外偏振度影响可以忽略;在理论分析基础上,有针对性的开展红外光谱偏振探测试验,试验数据分析与理论推导结论吻合。这表明：材料表面发射率的变化不影响目标表面的红外辐射偏振度。该研究成果有利于提高红外伪装探测的目标识别效率,可为进一步提高红外偏振成像系统的伪装目标探测提供新的途径和方法,如在伪装目标探测识别中,通过探测其表面的红外辐射偏振特性的改变实现伪装目标识别探测。     

火灾烟气中CO2和CO的释放特性研究

火灾产生的有害烟气是造成人员伤亡的主要原因。火灾科学研究中,研究者对CO_2和CO研究较多,但对它们的释放过程和机理研究很少。本文利用国际文献中评价材料毒性的原始实验数据,从烟气毒性物质释放的角度,分析了火灾烟气中 CO_2和 CO的释放特性,特别是分析了通风量的影响机理。在此基础上,进一步研究发展出一个与动态释放过程相关的烟气危害性评价体系,从而在释放规律与材料选择、实际火灾中人员的逃生过程之间建立起直接联系。     

概率神经网络在高功率微波探测数据处理中的应用

为了更好地处理高功率微波探测过程中产生的样本数据,在深入分析高功率微波特性参数的基础上,建立一个高功率微波器件特征参数库,并结合概率神经网络系统建立了一个高功率微波探测预测模型。通过部分学习样本和非学习样本进行预测,预测结果证明该模型能够基本再现原始数据,同时,对非样本数据有着较好的预测能力。这一数据处理方法在处理复杂样本、模式分类和判别过程中具有较高的实用性和实时性,能够在高功率微波探测数据的数据分类、结果预测等方面得到较好的应用。     

基于近红外光谱技术的多组分毒性气体检测研究

基于近红外波长调制光谱和谐波探测方法,并采用时分复用方案,建立了多组分毒性气体同时检测的实验系统。成功实现了对火灾产物中多组分毒性气体(包括CO,CO2,HCN)浓度的同时检测,并通过对三个通道的基频分量和二次谐波分量进行实时数据处理,计算得到浓度结果。三种毒性气体各组分的满量程体积分数分别为4×10-4,4×10-4,3×10-4,测量结果误差的最大值不超过满量程的0.73%。实验结果表明,该系统进行气体浓度检测具有较好的线性度,并满足检测精度要求。     

基于颜色模型和稀疏表示的图像型火焰探测

常用的图像型火焰探测算法是提取火焰在图像上表现出的单个特征信息或其有效组合作为识别的依据,需要大量的训练样本进行学习与参量优化,且识别率对特征选择的要求也很高.本文从火焰的整体特征考虑,提出了基于颜色模型和稀疏表示模型相结合的图像型火灾探测方法.首先在HIS空间建立颜色模型对火灾图像进行预处理提取出疑似区域,建立稀疏表...     

光声光谱技术在多组分气体浓度探测中的应用

CO和CH_4气体作为判断变压器运行状态的故障气体,对其浓度的探测在变压器维护中具有重要意义.为了准确探测变压器运行过程中产生的CH_4和CO气体浓度,本文利用光声光谱技术,设计了一套基于宽带光源的多组分气体探测系统,和共振型光声系统相比,该系统中所用的非共振型光声池体积小,易加工,池内各处信号强度相同,降低了对声学信号探测器的安装要求.系统的性能通过对CO和CH_4气体的探测进行评估.首先,从理论上分析了信号强度与调制频率呈反比,然后根据宽带光声系统在不同调制频率下的响应,确定系统的最佳调制频率为22 Hz.在最佳调制频率下,根据温度与待测气体光声信号的关系,对光声信号进行温度补偿,消除温度变化对光声信号的影响,进一步提高了系统的稳定性.最后,通过不同浓度的CH_4和CO气体对系统进行标定.实验表明,温度补偿前后,光声信号随温度的漂移分别为0.023 23V/℃和8.383 48×10~(-5) V/℃,通过对不同浓度CH_4和CO气体的探测,系统的线性度分别达到0.995和0.998 4.在一个大气压下,积分时间为1s时,宽带光声探测系统对CO和CH_4气体的探测极限浓度能够达到1μL/L.该系统成本低,线性度好,探测灵敏度符合国标对变压器维护过程中CO和CH_4气体的探测要求.     

基于STM32F电气火灾无线红外测温系统的设计与实现

为进一步提高电气火灾探测系统的探测精度及响应时间,在传统电气火灾探测系统的基础上,设计了基于STM32F的无线红外测温系统;该系统采用红外测温技术,探测出电气配电箱内各线路的温度信号;通过微处理器实时采集温度信号及A/D转换后的数据,进而进行数据处理和算法分析;并利用ZigBee无线网络技术组建检测网络,实现温度数据的可靠传输;实验结果表明:电气火灾无线红外测温系统精确度高达99.37%,灵敏度高,响应时间为20.5 s,能精确监测电气配电箱内温度,预防电气火灾的发生。     

Algorithm for detection of active fire zones using NOAA AVHRR data

Forest fires in tropical countries have become one of the important sources of Green House Gases and cause damage to the forest ecosystems. Information relating to the extent and location of active fire zones needs immediate attention to facilitate effective management and mitigation of forest fires. The spectral intervals that are most suitable for the detection of forest fires are 3.5–4.2 and 4.4–5.0 μm. The maximum thermal emission of fires is localized using middle infrared and thermal infrared bands for automatic recognition of fire active zones. The algorithm considers the thresholds for middle infrared on the basis of channel saturation to discriminate the fire zones from background warm surface and thermal infrared (10.3–11.3 μm) to ensure it to be free from cloud and water bodies. Advanced Very High Resolution Radiometer (AVHRR) data of 1 km resolution have been taken and we apply a spatial technique using background information to dynamically define thresholds appropriate for identifying fire pixels. The study suggests that satellite data provide a unique tool to detect and monitor active fire distribution.     

Application of a CO2 dial system for infrared detection of forest fire and reduction of false alarm

Forest fires can be the cause of serious environmental and economic damages. For this reason considerable effort has been directed toward forest protection and fire fighting. The means traditionally used for early fire detection mainly consist in human observers dispersed over forest regions. A significant improvement in early warning capabilities could be obtained by using automatic detection apparatus. In order to early detect small forest fires and minimize false alarms, the use of a lidar system and dial technique will be considered. A first evaluation of the lowest detectable concentration will be estimated by numerical simulation. The theoretical model will also be used to get the capability of the dial system to control wooded areas. Fixing the burning rate for several fuels, the maximum range of detection will be evaluated. Finally results of simulations will be reported. PACS 42.68.Wt; 89.60.Ec; 92.60.Mt; 92.60.Iq     

Detection of small forest fires by lidar

The possibility of detecting small forest fires with the help of a simple and cheap lidar operating at 0.532-μm wavelength up to distances of about 6.5 km is demonstrated. The values of the signal-to-noise ratio (SNR) achieved in the experiments are consistent with theoretical estimations obtained by computational modeling of the lidar detection process, including simulation of the smoke-plume shape and of the laser beam–plume interaction. This model was used to assess the potential of the lidar technique for fire surveillance in large forest areas. In particular, the upper limiting range for effective detection (SNR>5) of small localized fires in dry- and clear-weather conditions is estimated at 7–15 km depending on operation mode, burning rate, and observation geometry. Received: 29 August 2001 / Published online: 29 November 2001     

基于多光谱图像的烟雾检测

烟雾检测对于火灾早期防范非常重要,传统的智能视频和图像处理技术易受背景运动信息影响,抗干扰性差,且不容易区分森林水雾和燃烧产生的烟雾,森林防火误报率高。为此提出一种新的多光谱图像检测方法检测烟雾。采用多光谱成像系统,获取400至720 nm波段范围的烟雾、水雾光谱图像序列,对图像进行分层像素整合处理;利用欧氏距离度量不同分块光谱特征差异,获取动态区域光谱特征向量,根据目标与背景间光谱特征向量差异,提取烟雾、水雾区域。室内外试验结果表明：多光谱图像检测方法可用于烟雾检测,能够有效地检测并区分烟雾和水雾,与视频图像方法结合,可有效地用于森林火灾监测,降低森林火灾检测误报率。     

多光谱遥感影像煤火监测新方法

煤火在世界各地均有不同程度的发生,严重威胁生态环境,煤火燃烧释放大量的有毒有害气体,造成大气污染,同时煤火燃烧形成地下空洞,导致地表塌陷,直接威胁着矿区人员的生命安全。遥感技术的迅猛发展使大尺度反演与监测煤火温度变为可能。单窗算法是一个简单可行且精度较高的煤火温度反演方法,该方法需要两个大气参数进行温度反演,即大气水分含量和地表比辐射率。由于该算法需要卫星影像获取瞬间时的大气水分含量,而卫星过境瞬间的大气水分含量受多种因素影响难以获得;单窗算法对不同类型的地物均采用统一的地表比辐射率,这会导致反演温度不精确,误差较大。针对上述存在的问题,采用基于地面湿度参量建立起的大气可降水量与地面水汽压间的经验关系,计算大气水分含量,同时,采用NDVI阈值法计算不同地物类型的地表比辐射率,对单窗算法中的两个参数进行精确估计,从而改进提高该算法的精度及可操作性。将改进的算法应用于内蒙古乌达矿区,反演从1988年到2015年间研究区的煤火温度,提取每年研究区的温度异常区域,对比分析煤火区域分布、面积变化情况。本文提出的改进算法能够快速、高效的反演煤火温度,对掌握煤火异常区域变化情况提供技术支持,具有可操作性及现实意义。     

Nutrient cycling responses to fire frequency in the Kruger National Park (South Africa) as indicated by stable isotope analysis

Fires, which are an intrinsic feature of southern African ecosystems, produce biogenic and pyrogenic losses of nitrogen (N) from plants and soils. Because of the long history of fires in these savannas, it was hypothesized that N2 fixation by legumes balances the N losses caused by fires. In this study, the N2 fixation activity of woody legumes was estimated by analyzing foliar delta15N and proportional basal area of N2 fixing species along experimental fire gradients in the Kruger National Park (South Africa). In addition, soil carbon (C) and N pools, foliar phosphorus (P) and gross N mineralization and nitrification rates were measured, to indicate the effects of fires on nutrient stocks and the possible N cycling processes modified by fires. Although observations of increased soil C/N and mineralization rates in frequently burned plots support previous reports of N losses caused by fires, soil %N did not decrease with increasing fire frequency (except in 1 plot), suggesting that N losses are replenished in burned areas. However, relative abundance and N2 fixation of woody legumes decreased with fire frequency in two of the three fire gradients analyzed, suggesting that woody legume N2 fixation is not the mechanism that balances N losses. The relatively constant %N along fire gradients suggests that these ecosystems have other mechanisms to balance the N lost by fires, which could include inputs by atmospheric deposition and N2 fixation by forbs, grasses and soil cyanobacteria. Soil isotopic signatures have been previously used to infer patterns of fire history. However, the lack of a relationship between soil delta15N and fire frequency found in this study indicates that the effects of fires on ecosystem delta15N are unpredictable. Similar soil delta15N along fire gradients may reflect the contrasting effects of increased N gaseous emissions (which increases delta15N) and N2 fixation other than that associated with woody legumes (which lowers delta15N) on isotopic signatures.     

基于MODIS数据的森林火险时空分异规律研究

森林火灾严重危害生态环境,引起了全球的高度重视。将从MODIS(MODerate-resolution imaging spectroradiometer)中提取的活动火点与历史火烧痕迹进行比较研究,发现MOD14A1(火掩膜数据产品a daily Level 3 1-km fire hot spot product)中提取的8+9波段适合消防监测,与现场勘察数据相比较吻合度高达0.83。使用MOD14A1中8+9波段结合相关数据对这个区域的长达11年(2000—2010年)的森林火灾发生的时间和空间分析,结果表明：火灾发生频率最多的是春季,秋季次之,夏天概率最低,除非干旱。通过对研究区域黑龙江省分析,针叶林和温带针阔混交林过火面积所占比例分别为53.68%,44%,草原区过火面积较小为2.32%。大兴安岭是主要的燃烧区域,面积达到64.74%,小兴安岭约为23.49%,而其他区域面积不超过5%。且火灾发生的较大部分森林有个平缓的斜坡(≤5°),大部分处于中海拔(200 m≤H≤500 m)。因此,通过卫星遥感对森林火区区域的时间序列分析,阐明火灾活动规律和气候、地形、植被类型的相互关系,有助于预测火灾区域危险性等级。     

Nutrient cycling responses to fire frequency in the Kruger National Park (South Africa) as indicated by Stable Isotope analysis

Fires, which are an intrinsic feature of southern African ecosystems, produce biogenic and pyrogenic losses of nitrogen (N) from plants and soils. Because of the long history of fires in these savannas, it was hypothesized that N 2 fixation by legumes balances the N losses caused by fires. In this study, the N 2 fixation activity of woody legumes was estimated by analyzing foliar i 15 N and proportional basal area of N 2 fixing species along experimental fire gradients in the Kruger National Park (South Africa). In addition, soil carbon (C) and N pools, foliar phosphorus (P) and gross N mineralization and nitrification rates were measured, to indicate the effects of fires on nutrient stocks and the possible N cycling processes modified by fires. Although observations of increased soil C/N and mineralization rates in frequently burned plots support previous reports of N losses caused by fires, soil %N did not decrease with increasing fire frequency (except in 1 plot), suggesting that N losses are replenished in burned areas. However, relative abundance and N 2 fixation of woody legumes decreased with fire frequency in two of the three fire gradients analyzed, suggesting that woody legume N 2 fixation is not the mechanism that balances N losses. The relatively constant %N along fire gradients suggests that these ecosystems have other mechanisms to balance the N lost by fires, which could include inputs by atmospheric deposition and N 2 fixation by forbs, grasses and soil cyanobacteria. Soil isotopic signatures have been previously used to infer patterns of fire history. However, the lack of a relationship between soil i 15 N and fire frequency found in this study indicates that the effects of fires on ecosystem i 15 N are unpredictable. Similar soil i 15 N along fire gradients may reflect the contrasting effects of increased N gaseous emissions (which increases i 15 N) and N 2 fixation other than that associated with woody legumes (which lowers i 15 N) on isotopic signatures.     

Towards large eddy simulations of flame extinction and carbon monoxide emission in compartment fires

The general objective of this research is to adapt current combustion modeling capabilities used in computational fluid dynamics solvers to the treatment of under-ventilated compartment fires. More specifically, we consider in the present study two models proposed to describe: diffusion flame extinction due to air vitiation; and the emission of carbon monoxide (CO) and unburnt hydrocarbon (HC) mass in a compartment fire. The flame extinction model is based on a flammability diagram parametrized in terms of vitiated air properties. The CO/HC mass model is based on: a transport equation for fuel mass; a comparison of this fuel mass to a Burke–Schumann chemical-equilibrium expression; and an interpretation of the difference as a measure of incomplete combustion. Both models are implemented into a large eddy simulation solver developed by the National Institute of Standards and Technology, USA. The models performance is tested via detailed comparisons with an experimental database corresponding to reduced-scale compartment fires. The study considers two cases that correspond to different values of the fire room global equivalence ratio and are representative of strikingly different flame behaviors. The comparative tests serve to evaluate the general ability of the models to describe the transition from extinction-free conditions to conditions in which the flame experiences partial or total quenching, as well as the transition from fire regimes with no or little CO emission to regimes that emit hazardous levels.