基于主成分分析和集成神经网络的发动机故障诊断模型研究

针对发动机废气排放参数和故障之间复杂的非线性关系，提出了一种基于主成分分析和集成神经网络技术的发动机故障诊断分析模型。该模型首先运用主成分分析方法降低故障诊断样本的输入维数，然后按发动机不同运转状态将样本分组，并用于子网络训练；故障诊断时，各子网络分别诊断出相应的结果，最后采用投票法融合各输出结果。试验结果表明，这种模型能有效简化训练样本和样本属性参数，优化网络结构，其诊断精度及学习能力优于单一神经网络诊断模型，能较好地解决网络规模大、训练速度慢、诊断精度低等缺点。     

秸秆还田对灌溉玉米田土壤反硝化及N2O排放的影响

运用乙炔抑制技术研究了不同施氮水平下秸秆还田对灌溉玉米田土壤反硝化反应和氧化亚氮(N2O)排放的影响。结果表明,土壤反硝化速率及N2O的排放受氮肥施用、秸秆处理方式及其交互作用的显著影响。与秸秆燃烧相比,不施氮或低施氮水平时,秸秆还田可刺激培养初期反硝化反应速率及N2O排放,增加培养期间N2O平均排放通量;高施氮水平时,秸秆还田可降低反硝化反应速率及反硝化过程中的N2O排放。秸秆还田可降低反硝化中N2O/N2的比例。     

基于响应面法的离心泵作透平水力和声学性能优化

为综合优化离心泵作透平的水力和声学性能,建立了一种基于响应面的离心泵作透平水力和声学性能多目标优化方法。首先在对比分析叶轮几何参数对透平水力和噪声影响的基础上,根据敏感度筛选出对噪声影响显著的关键参数;进而应用响应面方法构造显著变量与多目标函数的响应面多元回归模型,分析影响水力效率与噪声的参数间交互作用;最终以水力效率不降低和总声压级最小为响应目标,兼顾性能与噪声确定最优参数组合,即叶片进口安放角为19.5°,叶片出口安放角为20°,叶片出口宽度为16 mm,叶片包角为92°,叶轮进口直径为101 mm,叶片数为12。对某离心泵作透平多目标优化结果表明,叶轮进口直径、叶片出口宽度、叶片数及叶片包角对内场噪声总声压级影响显著;响应面模型能够反映参数与响应值之间的相关性;经试验验证优化后透平水力效率平均提高了1.98个百分点,总声压级降低了4.95 d BA,表明采用的响应面法能够在不影响透平原有水力性能的前提下改善声学性能。     

稻田温度与甲烷排放通量关系的研究

稻田温度与CH4 排放通量有密切关系 ,通过灰关联分析发现稻田 5cm深处温度与CH4 排放通量关系最密切 ,水稻抽穗期CH4 排放通量达极大值 ,而完熟期CH4 排放通量达极小值。温度对CH4 排放通量的增效应明显 ,减效应较弱。稻田温度与CH4排放通量的关系呈S型曲线相关关系。     

Estimates of denitrification in soil by remote sensing of thermal infrared emission at different moisture levels

Remote sensing techniques may be one way to narrow the range of uncertainty in extrapolating N₂ emissions from small-scale to large-scale terrestrial ecosystems. In the present work we investigated the correlations between denitrification activity, soil moisture, and soil thermal infrared emissions. A field experiment was performed on two different agricultural soils, one loam and one silty clay. The results indicated that thermal infrared emissions can only be used to estimate the denitrification rate in soil within a limited range of soil moisture levels. Estimates of denitrification activity based on soil texture and moisture are, however, very likely to be a fruitful approach to generating large-scale N fluxes.     

Effect of fertilizer application on methane emission/production in the paddy soils of Taiwan

Methane production in three types of rice paddy soil was investigated under greenhouse conditions. The amount of methane produced during the first crop season (March to July) was 2–6 times higher than that in the second crop season (August to December). Application of organic fertilizer hastened the drop in redox potential and increased methane production and emission. Methane production also increased with the depth of soil with high values in soil samples from 18 to 30cm depth. Methane production in the first crop season was 18.0, 54.3 and 49.4mgcm^–3 for 6tha^–1 straw application for Linkou, Tzawchyau and Jiaushi soils, respectively. The value was 33.4mgcm^–3 for the second crop season in Jiaushi soil. Methane emission was high during the flowering and maturity stages in the first crop season and the values were high during the tillering and flowering stages in the second crop season. Methane emission was high in Tzawchyau and Jiaushi soils in the first crop season. Methane emission rate reached a maximum from 12 noon to 3p.m. due to high temperature and a minimum at 3 to 6a.m. in both planted and unplanted soils. Received: 17 September 1996     

稻季施肥管理措施对后续麦季N2O排放的影响

2003─2004年选用江苏省宜兴市稻-麦轮作试验田,研究了水稻生长季秸杆(0和3.75×103kg/hm2两个水平)和N肥施用(N0、200和270kg/hm23个水平)对后续麦季N2O排放的影响。结果表明:稻季秸杆施用显著减少了后续麦季N2O的排放,这些减少量主要体现在小麦播种-返青期,方差分析达显著水平(P<0.05)。稻季施用N肥,后续麦季N2O排放减少,但N200和270kg/hm2N肥施用水平的处理间无显著差异。麦季土壤水分情况与N2O排放通量存在显著正相关(P<0.05)。     

Fluxes of methane from rice fields and potential for mitigation

Abstract. Methane (CH₄) is an important greenhouse gas. Flooded rice fields (paddies) are a significant source of atmospheric CH₄; estimates of the annual emission from paddies range from less than 20 to 100 million Tg, with best estimates of 50 × 20 Tg. The emission is the net result of opposing bacterial processes: production in anaerobic microenvironments, and consumption and oxidation in aerobic microenvironments, both of which occur sequentially and concurrently in flooded rice soils. With current technologies, CH₄ emission from rice fields will increase as production increases. Over the next 25 years rice production will have to increase by 65% from the present 460 Mt/y to 760 Mt/y in 2020. The current understanding of the processes controlling CH₄ fluxes, rice growth and rice production is sufficient to develop mitigation technologies. Promising candidates are changes in water management, rice cultivars, fertilization, and cultural practices. A significant reduction of CH₄ emission from rice fields, at the same time that rice production and productivity increase at the farm level, is feasible, although the regions where particular practices can be applied, and the trade-offs that are possible, have still to be identified.     

中国农村可再生能源技术应用对温室气体减排贡献的研究

该文以国内外温室气体减排和能源领域的相关研究为基础，根据中国农村可再生能源技术特点，研究提出了适于农村可再生能源特点的燃料燃烧温室气体排放系数和粪便管理CH₄排放系数，以及CO₂和CH₄减排量的计算公式，并分析了农村可再生能源技术对减排温室气体的贡献。     

妊娠猪舍氨气及氧化亚氮浓度测定与排放通量的估算

在北京选择一典型猪场，对不同季节妊娠舍的氨气及氧化亚氮浓度进行了为期1年的试验测定，并根据二氧化碳平衡原理，对猪场不同生长阶段的妊娠猪含氮气体的排放通量进行了估算。结果表明：冬季2004年11月氨气和氧化亚氮的平均浓度分别为11.64±4.36，（1.17±0.2）mg／m³，夏季7月舍内氨气和氧化亚氮的平均浓度分别为3.31±2.67，（0.6±0.02）mg／m³；妊娠猪饲养期间的氨气排放通量为每头（185.2～500.8）mg／h，氧化亚氮为（3.85～35.93）mg／h。