基于机器学习的热释放速率实时预测方法

火灾态势的实时感知对提升消防救援指挥与决策水平有非常重要的意义。为实现火源热释放速率的实时预测,通过丙烷气体燃烧实验获取火焰体积和高度的实时数据,选取MATLAB中精细高斯SVM做分类训练,建立了基于图像处理的热释放速率预测模型。结果表明,训练完成后的模型可以根据3 s内的火焰体积/高度数据预测得到实时的燃料流量,获取相应的火源热释放速率。该预测模型在已知数据集和未知数据集上都表现出了较高的准确性,为火灾态势发展预测提供了理论和技术基础,有很好的应用前景。     

炭膜曝气生物膜反应器硝化作用及其微生物群落结构分析

采用炭膜曝气生物膜反应器处理无机含氮废水,通过改变进水氨氮浓度和水力停留时间,研究了反应器硝化性能、氧利用情况以及氨氮去除负荷,并对生物膜表面特性和硝化菌优势菌种进行分析.结果表明, 在膜内气压0.017 MPa,进水NH⁺₄-N 50 mg/L,HRT为8 h条件下,NH⁺₄-N去除率达到96%,出水NO^-₂-N平均为17 mg/L,一定程度上实现了短程硝化,炭膜所供给氧气被生物膜全部消耗;系统比表面氨氮最大去除速率为9.7 g/(m²·d),炭膜表面有限的生物量制约了去除速率的进一步提高; 荧光原位杂交技术分析揭示生物膜内亚硝化单胞菌（Nitrosomonas）和亚硝化螺菌（Nitrosospira）为亚硝化细菌优势菌种,分别占细菌总菌数的19%和21%,硝化螺菌（Nitrospira）为硝化细菌优势菌种,占总菌数的20%,未检测到硝化杆菌（Nitrobacter）的存在.     

超声波/零价铁协同降解氯苯类化合物的QSPR研究

在同一反应条件下测得超声波/零价铁(US/Fe)协同降解五种氯苯类化合物符合准一级反应,降解速率常数(k)呈现一定的规律性:1,2,4-TCB>M-DCB>P-DCB>O-DCBCB>CB。采用偏最小二乘法(PLS)建立氯苯类化合物定量结构-性质相关(QSPR)的数学模型,模型的主成分分析结果表明:在US/Fe体系中,影响氯苯类有机物降解的主要因素分别是物质的氯取代程度,分子极化率及亨利常数。     

PLS and PCR Methods in the Assessment of Coastal Water Quality

Partial least squares regression analysis (PLS) and principal component regression analysis (PCR) were examined asmethodological procedures for assessing the quality of coastalwaters in a tourist area. Four variables related to the trophicstate of waters, namely nitrate, nitrite, ammonia and phosphate were analyzed. The models resulting from PLS and PCR were verysimilar. Both defined three groups of water masses characterizedby different nutrient loadings. These groups were in accordancewith those obtained by numerical classification. The PLS methodwas selected as the optimal model, on the basis of its lowerprediction errors (lower Press and Rmsd values). For managementpurposes, this statistical model allows mesotrophic conditions,reflecting some nutrient enrichment over background conditions,to be characterized and the successful diagnosis of additionalsamples within this context.     

The Seasonal Variation of the Concentrations of Ozone, Sulfur Dioxide, and Nitrogen Oxides in Two Nigerian Cities

A combination of multivariate statistical methods including factor analysis, principal component analysis, principal component regression, and multiple linear regression (MLR) were employed to evaluate the influence of seasons on the concentrations of ozone, sulfur (IV) oxide, and oxides of nitrogen in ambient air of Nigerian cities of Lagos and Ilorin. The former city is located in the coastal area, and it is highly congested with a high intensity of marine, vehicular, and industrial activities, and the latter city is a medium size town, located in the central guinea savannah zone of Nigeria. Samples were collected using a high-volume sampler from near the ground at various sites of diverse human and industrial activities, during wet and dry seasons from 2003 to 2006. The PCA reveals three distinct groupings during the day for all data, which is a reflection of different factors contributing to the atmospheric chemistry of these cities. The predicted ozone concentration values by MLR agree fairly well with the measured data. The dependence of ozone on meteorological parameters including relative humidity, air temperature, and sun exposure and the precursor pollutants depends on weather and the anthropogenic activities. The results for the two cities indicate that reduction in the level of NO₂ is accompanied by an increase in the level of ozone, suggesting the interconversion between the two via photochemical activity.     

不同冬夏寄主棉蚜种群重复序列引物DNA多态性分析

应用重复序列引物ＰＣＲ技术研究了中国不同冬夏寄主上棉蚜种群的ＤＮＡ多态性,通过筛选出的３种适用于棉蚜的重复序列引物,用它们的扩增结果进行相似性指数Ｉｓ和Ｎｅｉ’ｓ遗传距离Ｄ的计算,并根据遗传距离对所研究的棉蚜种群做聚类分析,结果表明,用遗传距离指标可以将冬夏寄主上的棉蚜分开,并且在不同的冬寄主之间,较适宜越冬的不同寄主上的棉蚜呈现出更大的相似性。在实验数据的分析过程中,还发现遗传距离指标比相似性指     

Nutrient depletion is the main limiting factor in the crude oil bioaugmentation process

The biodegradation was considered as the prime mechanism of crude oil degradation. To validate the efficacy and survival of the crude oil-degrading strain in a bioremediation process, the enhanced green fluorescent protein gene (egfp) was introduced into Acinetobacter sp. HC8–3S. In this study, an oil-contaminated sediment microcosm was conducted to investigate the temporal dynamics of the physicochemical characterization and microbial community in response to bacterium amendment. The introduced strains were able to survive, flourish and degrade crude oil quickly in the early stage of the bioremediation. However, the high abundance cannot be maintained due to the ammonium (NH₄⁺-N) and phosphorus (PO₄^3?-P) contents decreased rapidly after 15 days of remediation. The sediment microbial community changed considerably and reached relatively stable after nutrient depletion. Therefore, the addition of crude oil and degrading cells did not show a long-time impact on the original microbial communities, and sufficient nitrogen and phosphorus nutrients ensures the survive and activity of degrader. Our studies expand the understanding of the crude oil degradative processes, which will help to develop more rational bioremediation strategies.     

城市生态资源定量评估

城市生态资源定量评估对区域可持续发展具有重要意义. 以城市特色明显的深圳为例,在分析其生态系统特征基础上,提出了深圳市生态资源测算指标体系和分级标准. 该指标体系包括植被覆盖指数、水面覆盖指数、建设用地指数、未利用地指数和污染物排放指数五大类二级指标,以及林地等17个三级指标. 运用所构建的指标体系对深圳市2007年生态资源状况进行了评估,结果表明,罗湖、盐田、龙岗3个区生态资源优秀,福田、宝安、光明3个区生态资源良好,南山生态资源中等,全市生态资源总体表现良好. 该研究结果与深圳市资源环境实际状况基本一致.      

取代苯酚类化合物抑藻活性的CoMFA模型

基于比较分子力场分析(CoMFA)方法建立20种取代苯酚类化合物抑藻活性(pI)的三维定量构效关系(3D-QSAR)。训练集中16个化合物用于建立预测模型,测试集5个化合物(含模板分子)作为模型验证。已建立的Co MFA模型的交叉验证系数(Q2)、非交叉验证系数(R2)分别为0.915、0.963,说明所建模型具有较强的稳定性和良好的预测能力。该模型中立体场、静电场贡献率依次为48.4%、51.6%,表明影响抑藻活性(pI)的主要因素是取代基的电荷分布,其次是取代基的疏水性和空间位阻。