陶瓷原料分类的Bayes判别分析法

将Bayes判别分析法应用于陶瓷原料分类中,建立了陶瓷原料分类识别的Bayes判别分析模型.模型选用陶瓷原料的9种化学成分作为分类识别的判别因子;将陶瓷原料分为4个类别并作为Bayes判别分析的4个正态总体;以陶瓷原料样本实测数据作为训练样本,建立Bayes线性判别函数;以Bayes线性判别函数计算待判样品的Bayes判别函数值,以最大值对应的总体作为样品所归属的总体;最后以回代法对判别准则进行评价以检验模型的优良性.实例研究表明,Bayes判别分析模型回判估计误判率为0,因而是陶瓷原料分类识别的一种有效方法,可在生产实际中应用和推广.     

基于光谱分类模型的保险杠物证无损研究

采用中红外光谱结合化学计量学的方法对车用保险杠碎片进行鉴别,分别对52个车用保险杠碎片样本的全波段光谱数据、指纹区光谱数据和主成分分析降维后的光谱数据建立Fisher判别分析和K近邻算法2种分类模型,并对分类结果进行比较。结果表明,主成分分析提取特征变量后构建的分类模型,分类的准确率更高,对聚丙烯（PP）、PP/滑石粉、PP/滑石粉/碳酸钙（CaCO₃）3种类型的样本分类准确率达到92.3 %,对PP/滑石粉类型中的10种品牌样本分类准确率达到88.9 %,分类结果理想;在构建的2种分类模型中,Fisher判别分析模型的分类率远高于K近邻算法模型,分析认为K近邻算法模型受到样本不均衡的影响;中红外光谱结合化学计量学可以实现对车用保险杠碎片的准确区分,且满足快速、无损的检验要求。     

文本分类中特征选择方法的研究与比较

介绍了进行文本分类的关键技术,并着重介绍了常用的文本特征提取方法。选取支持向量机方法作为文本分类器方法,选取不同特征提取方法应用于文本分类,通过实验,比较和分析了由不同的提取方法所构成的分类器的分类性能,确定了信息增益（IG）法和文本证据权（W ET）为两种性能优异的特征提取方法。该结论可为分类性能进一步的优化研究奠定理论和实践基础。     

化学模式识别及其在水质分类中的应用

本文简述了化学模式识别方法的原理和特点．着重论述了线性学习机方法。将该法用于南通和盐城两地的地表水水质分类．结果表明本文建立的方法识别率和预测准确率均达到100％。由于线性学习机具有自学习和自适应能力．可以较好地消除主观因素对分类的影响。     

文本分类中的特征提取方法的研究

文本的特征提取是文本分类过程中的一个重要环节,它的好坏将直接影响文本分类的准确率。该文介绍了词条的统计方法(CHI)、词条与类别的互信息(MI)、信息增益(IG)、词条的期望交叉熵(CE)等文本特征提取方法,并对其取词策略进行了改进。     

储层分类与钻井目标确定

储层分类标准的确定，是储量动用状况评价的基础，是油田开发潜力分析、确定钻井目标，进一步增加可采储量的关键。本文基于大量资料分析基础上，提出利用相对产能作为储层分类标准。相对产能不仅能反应井纵向的差异，同时可以显示储层平面上的变化情况，以此数据作为确定储层分类标准的依据是可行的。定量标准的确定，较大程度的简化了潜力分析的程度，为油田开发分析研究，制定下步挖潜方案及钻井目标，提供了有力工具。     

陶瓷原料分类的距离判别分析法

将距离判别分析法应用于陶瓷原料分类中,建立陶瓷原料分类的距离判别分析模型.模型选用陶瓷原料的9种化学成分作为距离判别分析模型的判别因子.将陶瓷原料分为4个类别并作为距离判别分析的4个正态总体,以陶瓷原料样本实测数据作为训练样本,建立相应线性判别函数对待判样本进行分类.研究结果表明,距离判别分析模型分类性能良好,回判估计的误判率为0,可以在生产实际中推广应用.     

利用压电传感器基于GAF-ResNet的管道焊缝缺陷分类

针对管道焊缝缺陷分类难度大的问题,提出了利用压电传感器数据,结合格拉姆角场（Gramian AngularField,GAF）和残差神经网络（ResNet）的焊缝缺陷分类方法。先采用GAF原理将一维时间序列数据转化为二维图像,将转化后的二维图像数据集输入,训练最优二维残差神经网络模型用于焊缝缺陷分类。实验中管道焊缝预制了10个缺陷（5种类型）,使用导波和超声技术分别对焊缝中1-5号缺陷进行检测,分析Precision（精确率）、Recall（召回率）、F1-score(F1评分)三个指标,证实了基于GAF-ResNet方法的可行性,同时6-10号缺陷验证了该方法的可靠性和普适性。     

红外指纹区光谱结合多阶导数融合技术无损分类鞋底物证

为了提高检验效率,降低检验鉴定成本,实现对鞋底的快速无损分类。采用傅里叶变换红外指纹光谱及其多阶导数光谱对5类不同品牌共计50个样本的鞋底进行分析,并构建Bayes判别和支持向量机2种分类模型。结果表明,在鞋底鉴别过程中,基于原始数据、一阶导数数据和二阶导数数据建立的融合模型,初级融合模型的区分效果优于单一模型和中级融合模型,总体分类准确率能达到80 %以上。而基于初级模型进行的成分特征提取中,BDA结合原始数据结合一阶导数模型是最好的,总体分类准确率达到92 %。红外指纹光谱结合一阶求导、二阶求导构建不同的融合模型进行区分对比,选择最为有效的融合模型可实现对日常皮鞋、运动鞋鞋底快速的无损鉴别,对今后的治安工作作具有借鉴意义,不仅缩小排查范围,也为案件的快速侦破提供了一种新的方式。     

陶瓷原料分类的人工神经网络模型

抽取陶瓷原料特征参数,建立3层BP人工神经网络模型,据此对陶瓷原料进行模式识别和分类.研究结果表明:经该模型计算所得的分类结果和实际情况相吻合.该方法有助于在配方过程中遴选陶瓷原料,值得推广.     

Helix stabilizing factors and stabilization of thermophilic proteins: an X-ray based study

We have compared the X-ray structures of 13 thermophilic proteins with their mesophilic homologues, in order to bring out differences in the stability of helices. The energy terms of a helix-coil transition algorithm were used to evaluate helix stability. Helices of thermophilic proteins are more stable than the mesophilic homologues in 69% of cases. This is due mainly to intrinsic helical propensities of amino acids, whereas minor effects are linked to main chain H-bonds, side chain-side chain interactions, capping motifs and charge-dipole effects. Furthermore, the frequency of 10 helix stabilizing factors recognized by appropriate sequence patterns was evaluated. The only factor occurring significantly in the thermostable proteins was the lack of beta branched residues. Other factors do not show a definite trend, although their occurrence in proteins is believed to be important for stability. This is discussed in the light of protein engineering applications.      

Factors enhancing protein thermostability

Several sequence and structural factors have been proposed tocontribute toward greater stability of thermophilic proteins.Here we present a statistical examination of structural andsequence parameters in representatives of 18 non-redundant familiesof thermophilic and mesophilic proteins. Our aim was to lookfor systematic differences among thermophilic and mesophilicproteins across the families. We observe that both thermophilicand mesophilic proteins have similar hydrophobicities, compactness,oligomeric states, polar and non-polar contribution to surfaceareas, main-chain and side-chain hydrogen bonds. Insertions/deletionsand proline substitutions do not show consistent trends betweenthe thermophilic and mesophilic members of the families. Onthe other hand, salt bridges and side chain–side chainhydrogen bonds increase in the majority of the thermophilicproteins. Additionally, comparisons of the sequences of thethermophile–mesophile homologous protein pairs indicatethat Arg and Tyr are significantly more frequent, while Cysand Ser are less frequent in thermophilic proteins. Thermophilesboth have a larger fraction of their residues in the     

Nitrogen requirement for the mesophilic and thermophilic upflow anaerobic filters of a simulated papermill wastewater

This study determined the minimum nitrogen concentration at which the mesophilic and thermophilic upflow anaerobic filters treating a simulated papermill wastewater could operate efficiently. For this purpose, the chemical oxygen demand (COD) to nitrogen ratio in both digesters was increased in five consecutive steps, from 94: 5 to 363: 5. The maximum COD to nitrogen ratios that provided satisfactory operation were 283: 5 (30mg N/L) and 363: 5 (23 mg N/L) for the mesophilic and thermophilic reactors, respectively. The nitrogen consumption of bacteria in the thermophilic digester was lower than that in the mesophilic digester. The operational efficiency of the thermophilic digester was higher than that of the mesophilic one regardless of the COD to nitrogen ratio.     

沼液余热回收对高温发酵沼气工程净产气率的影响

以瑞典Alviksgården养猪场高温发酵沼气工程为分析对象,通过传热模型计算发酵罐加热负荷,通过与江苏金坛永康农牧养猪场同等规模的中温发酵沼气工程对比,讨论高温发酵的容积产气率和余热回收对沼气工程净产气率的影响。结果表明虽然瑞典Alviksgården养猪场高温发酵沼气工程的加热能耗是江苏金坛永康农牧养猪场中温发酵沼气工程的2.1倍,但由于产气速率达到2.3 m³·m^-3·d^-1,增产的沼气量远远大于增温的能耗;若进一步在瑞典Alviksgården养猪场沼气工程中增加余热回收,可将沼气净产气率从82%提高至90%。相对于提高容积产气率,采用余热回收技术对降低高温发酵沼气工程的增温能耗、提高净产气率效果更加显著。     

Contribution of engineered electrostatic interactions to the stability of cytosolic malate dehydrogenase

Protein engineering is a promising tool to obtain stable proteins.Comparison between homologous thermophilic and mesophilic enzymesfrom a given structural family can reveal structural featuresresponsible for the enhanced stability of thermophilic proteins.Structures from pig heart cytosolic and Thermus flavus malatedehydrogenases (cMDH, Tf MDH), two proteins showing a 55% sequencehomology, were compared with the aim of increasing cMDH stabilityusing features from the Thermus flavus enzyme. Three potentialsalt bridges from Tf MDH were selected on the basis of theirlocation in the protein (surface R176-D200, inter-subunit E57–K168and intrasubunit R149–E275) and implemented on cMDH usingsite-directed mutagenesis. Mutants containing E275 were notproduced in any detectable amount, which shows that the energypenalty of introducing a charge imbalance in a region that wasnot exposed to solvent was too unfavourable to allow properfolding of the protein. The salt bridge R149–E275, ifformed, would not enhance stability enough to overcome thiseffect. The remaining mutants were expressed and active andno differences from wild-type other than stability were found.Of the mutants assayed, Q57E/L168K led to a stability increaseof 0.4 kcal/mol, as determined by either guanidinium chloridedenaturalization or thermal inactivation experiments. This resultsin a 15°C shift in the optimal temperature, thus confirmingthat the inter-subunit salt bridge initially present in theT.flavus enzyme was formed in the cMDH structure and that theextra energy obtained is transformed into an increase in proteinstability. These results indicate that the use of structuralfeatures of thermophilic enzymes, revealed by a detailed comparisonof three-dimensional structures, is a valid strategy to improvethe stability of mesophilic malate dehydrogenases.     

Aromatic clusters: a determinant of thermal stability of thermophilic proteins

A number of factors have been elucidated as responsible forthe thermal stability of thermophilic proteins. However, thecontribution of aromatic interactions to thermal stability hasnot been systematically studied. In the present investigationwe used a graph spectral method to identify aromatic clustersin a dataset of 24 protein families for which the crystal structuresof both the thermophilic and their mesophilic homologues areknown. Our analysis shows a presence of additional aromaticclusters or enlarged aromatic networks in 17 different thermophilicprotein families, which are absent in the corresponding mesophilichomologue. The additional aromatic clusters identified in thethermophiles are smaller in size and are largely found on theprotein surface. The aromatic clusters are found to be relativelyrigid regions of the surface and often the additional aromaticcluster is located close to the active site of the thermophilicenzyme. The residues in the additional aromatic clusters arepreferably mutated to Leu, Ser or Ile in the mesophilic homologue.An analysis of the packing geometry of the pairwise aromaticinteraction in the additional aromatic clusters shows a preferencefor a T-shaped orthogonal packing geometry. The present studyalso provides new insights for protein engineers to design thermostableand thermophilic proteins.     

Molecular dynamics studies on the thermostability of family 11 xylanases

Twelve members of the family 11 xylanases, including both mesophilic and thermophilic proteins, were studied using molecular dynamics (MD). Simulations of xylanases were carried out in an explicit water environment at four different temperatures, 300, 400, 500 and 600 K. A difference in thermotolerance between mesophilic and thermophilic xylanases became clear: thermophilic xylanases endured heat in higher simulation temperatures better than mesophilic ones. The unfolding pathways seemed to be similar for all simulations regardless of the protein. The unfolding initiates at the N-terminal region or alternatively from the alpha-helix region and proceeds to the 'finger region'. Unfolding of these regions led to denaturated structures within the 4.5 ns simulation at 600 K. The results are in agreement with experimental mutant studies. The results show clearly that the stability of the protein is not evenly distributed over the whole structure. The MD analysis suggests regions in the protein structure which are more unstable and thus potential targets for mutation experiments to improve thermostability.     

Sequence and homology model of 3-isopropylmalate dehydrogenase from the psychrotrophic bacterium Vibrio sp. I5 suggest reasons for thermal instability

The leuB gene from the psychrotrophic strain Vibrio sp. I5 has been cloned and sequenced. The gene codes for 3-isopropylmalate dehydrogenase, a 360-residue, dimeric enzyme involved in the biosynthesis of leucine. Three recently solved homologous isopropylmalate dehydrogenase (IPMDH) crystal structures from thermophilic and mesophilic organisms have been used to build a homology model for the psychrotrophic IPMDH and to deduce the possible structural reasons for its decreased thermostability. According to our model the psychrotrophic IPMDH contains fewer stabilizing interactions than its mesophilic and thermophilic counterparts. Elements that have been identified as destabilizing in the comparison of the psychrotrophic, mesophilic and thermophilic IPMDHs are a smaller number of salt-bridges, a reduction in aromatic-aromatic interactions, fewer proline residues and longer surface loops. In addition, there are a number of substitutions of otherwise strictly conserved residues that can be linked to thermostability.      

畜禽粪便高温发酵与秸秆热化学处理工艺的耦合

我国规模化养殖场沼气工程普遍采用常温和中温发酵技术, 产气速率低导致发酵罐体积大和投资高, 是制约其快速发展的关键因素。另一方面, 秸秆作为农村常见低劣生物质常常被随意焚烧而导致严重的空气污染。为此, 本文提出将畜禽粪便高温发酵与秸秆热化学燃烧耦合的新思路, 通过秸秆燃烧为高温发酵过程提供热量, 将常温和中温发酵转变为高温发酵过程, 提高畜禽粪便厌氧发酵的处理速率, 同时避免秸秆的浪费。对耦合新工艺的计算结果表明, 对于万头猪场沼气工程, 发酵温度从30℃提高到55℃, 容积产气率由1.43 m³·m^-3·d^-1提高至3.40 m³·m^-3·d^-1, 发酵罐体积从1200 m³减小到500 m³;为维持高温发酵(55℃)所需热量, 年利用秸秆339 t。     

中温和高温厌氧消化的比较

厌氧消化目前已广泛应用于市政污泥、有机废水、城市有机垃圾、养殖粪便等有机废弃物的资源化处理,由于其具有在实现有机废弃物处理的同时产生甲烷燃料,剩余物还可以作为肥料使用的优点,全球各地已建立了许多大规模的厌氧消化工厂。温度是影响厌氧消化的重要因素,中温和高温厌氧消化是目前常采用的两种工艺,二者各有优劣,但是我国存在对高温厌氧消化的重视程度远远不及中温厌氧消化的问题。因此,本文首先对中温和高温厌氧消化及比较两种工艺的中英文文献进行了统计分析,然后重点从运行参数、有机物降解率、系统稳定性、产沼气性能、消化污泥特性（包括脱水性能、起泡潜力和病原菌含量）和微生物特性等运行性能、厌氧消化模型、经济效益和应用现状等方面对中温和高温厌氧消化进行了详细的比较,并对厌氧消化的应用和发展提出了建议。