复杂背景环境下基于SSD_MobileNet深度学习模型的火焰图像识别研究

针对现有的火灾火焰图像识别方法在光照和红花等类似火焰干扰的复杂环境下存在错检和漏检的问题,提出一种基于SSD_MobileNet的复杂环境火焰区域标记方法.首先,将深度卷积神经网络SSD300的基础卷积网络VGG16替换为MobileNet网络,应用深度可分离卷积,降低网络参数,进而构建一种火焰图像检测的SSD_Mob...     

我国大气VOCs监测现状及挑战

对我国开展大气VOCs监测的必要性、监测现状以及主要使用的监测方法进行了分析。目前,我国已初步建立了大气VOCs手工与自动监测网络,主要采用预浓缩-热脱附-气相色谱-质谱/氢火焰离子化检测器法进行PAMS组分和TO-15组分分析,利用高效液相色谱法进行13种醛、酮类组分分析。目前大气VOCs监测还存在数据质量、灵敏度有待提高,不同设备或方法监测结果一致性较差等问题。为此在VOCs监测过程中应进一步加强质量保证与质量控制,并尝试通过提高预浓缩装置除水和干扰物效率、提升进样量等多种手段提高监测灵敏度。     

流动注射在线离子交换预富集—火焰原子吸收法测定环境水样中痕量镍

研究了流动注射在线离子交换预富集与火焰原子吸收光谱法的联用技术,采用了双柱交替正向富集和反向洗脱的在线离子交换流路系统。在采样频率为３０次／ｈ下,灵敏度比普通火焰原子吸收法提高１２倍,应用于环境水样中痕量镍的监测,获得了满意结果。     

Volatile organic compounds at swine facilities: A critical review

Volatile organic compounds (VOCs) are regulated aerial pollutants that have environmental and health concerns. Swine operations produce and emit a complex mixture of VOCs with a wide range of molecular weights and a variety of physicochemical properties. Significant progress has been made in this area since the first experiment on VOCs at a swine facility in the early 1960s. A total of 47 research institutions in 15 North American, European, and Asian countries contributed to an increasing number of scientific publications. Nearly half of the research papers were published by U.S. institutions.Investigated major VOC sources included air inside swine barns, in headspaces of manure storages and composts, in open atmosphere above swine wastewater, and surrounding swine farms. They also included liquid swine manure and wastewater, and dusts inside and outside swine barns. Most of the sample analyses have been focusing on identification of VOC compounds and their relationship with odors. More than 500 VOCs have been identified. About 60% and 10% of the studies contributed to the quantification of VOC concentrations and emissions, respectively. The largest numbers of VOC compounds with reported concentrations in a single experimental study were 82 in air, 36 in manure, and 34 in dust samples.The relatively abundant VOC compounds that were quantified in at least two independent studies included acetic acid, butanoic acid (butyric acid), dimethyl disulfide, dimethyl sulfide, iso-valeric, p-cresol, propionic acid, skatole, trimethyl amine, and valeric acid in air. They included acetic acid, p-cresol, iso-butyric acid, butyric acid, indole, phenol, propionic acid, iso-valeric acid, and skatole in manure. In dust samples, they were acetic acid, propionic acid, butyric acid, valeric acid, p-cresol, hexanal, and decanal. Swine facility VOCs were preferentially bound to smaller-size dusts.Identification and quantification of VOCs were restricted by using instruments based on gas Chromatography (GC) and liquid chromatography (LC) with different detectors most of which require time-consuming procedures to obtain results. Various methodologies and technologies in sampling, sample preparation, and sample analysis have been used. Only four publications reported using GC based analyzers and PTR-MS (proton-transfer-reaction mass spectrometry) that allowed continuous VOC measurement. Because of this, the majority of experimental studies were only performed on limited numbers of air, manure, or dust samples. Many aerial VOCs had concentrations that were too low to be identified by the GC peaks.Although VOCs emitted from swine facilities have environmental concerns, only a few studies investigated VOC emission rates, which ranged from 3.0 to 176.5 mg d⁻¹ kg⁻¹ pig at swine finishing barns and from 2.3 to 45.2 g d⁻¹ m⁻² at manure storages. Similar to the other pollutants, spatial and temporal variations of aerial VOC concentrations and emissions existed and were significantly affected by manure management systems, barn structural designs, and ventilation rates.Scientific research in this area has been mainly driven by odor nuisance, instead of environment or health concerns. Compared with other aerial pollutants in animal agriculture, the current scientific knowledge about VOCs at swine facilities is still very limited and far from sufficient to develop reliable emission factors.     

Optimization of the ionization conditions for the trace analysis of PCDD/PCDF with ion trap MS/MS

Commercial ion trap mass spectrometer provides easy-to-operate MS/MS analysis for the determination of PCDD/PCDF. The limit of quantification is appropriately low (0.2 pg for 23478-P5CDF) because all the stages are performed in the trap and sample losses associated with the ion transportation are minimized. However, if excessive ions are injected into the trap, its electrical fields are distorted and an overall reduction in performance arises. Ionization condition is an important parameter as it affects the amount of the total ions produced. If the amount of interfering compounds are negligible, such as standard solution or cleaned-up sample, lower ionization condition (e.g. electron energy: 30 eV, emission current (EC): 150 μA) is preferable. On the contrary, in case excessive interfering ions are coexisting with PCDD/PCDF, such as crude extract or semi-cleaned-up sample, the ionization condition should be high (e.g. electron energy: 90 eV, EC: 350 μA) for the reproductive quantification.