探讨建立QR织物CAD组织库的新方法--按织物表面纹理效应选择织物组织

建立织物组织库是面向面料QR的智能化织物CAD系统的一个发展趋势，其中科学管理组织库是关键。通过分析国内外组织分类法的优缺点，以及信息时代生产技术提高、面料设计个性化发展对传统织物的组织分类法的冲击，阐明对织物组织重新分类的必要性。借鉴图书分类法，提出依据织物的实际外观效应、结合织物结构特征对组织进行分类的新方法。     

多数据源中文报表智能生成系统的研究

本文以多数据源的中文报表为研究对象 ,运用面向对象的方法 ,对其组成结构进行分析 ,提出了中文报表的生成四要素 ,并根据其数据来源的种类、限制条件、统计条件、各种综合条件以及各种相关的处理情况 ,从理论上提出报表单元格内容定义生成的解决方案     

分子马达定向运动的瞬态过程

本文研究了在一维三态周期跳跃模型下分子马达定向运动的瞬态过程，对于给定的任意初始分布，得出了与时间有关的几率分布的解析表达式，包括到达稳态之前的所有的瞬态过程，由此可获得马达在各个时刻的漂移速度和扩散系数，这是与对时间有关的两态模型的一上推广，同时还计算了马达到达稳态所需要的特征时间，并把结果同实验及其他计算结果进行了比较，与实验值符合的较好。     

东北红豆杉细胞悬浮培养研究

东北红豆杉（Taxus cuspidata sieb et zucc)幼茎诱导的愈伤组织经多次继代培养，选择生长旺盛，紫杉醇含量高的愈伤组织进行细胞悬浮培养，建立了东北红豆杉细胞悬浮系，同时，对影响愈伤组织进行细胞悬浮培养，建立了东北红豆杉细胞悬浮系，同时，对影响愈伤组织继代培养及细胞悬浮培养的条件进行了研究，确定了适合愈伤组织生长的条件以及影响细胞悬浮培养的主要因素。     

Microelement analysis of source of raw materials of terracotta warriors and horses of Qin’s Mausoleum in pits No. 1 and No. 2

In order to search for sources of the raw material of the Qin’s terracotta warriors and horses, we selected 20 samples of pottery warriors and pottery horses of the Qin’s Mausoleum from pit No. 1, 19 samples of pottery warriors and brick from pit No. 2, 20 samples of clays near the Qin’s Mausoleum and 2 samples of Yaozhou porcelain bodies. The 32 kinds of element content of each sample were measured by neutron activation analysis (NAA). These data were analyzed by fuzzy cluster analysis, the trend cluster analysis diagram was obtained. The results showed that samples of terracotta warriors and horses from pits No.1 and No. 2 were closely related to the loam layer near Qin’s Mausoleum, but their relations with the loess layers were not so close, and they were out of all relation to Yaozhou porcelain bodies. Thus, it could be deduced that the Li Mountain clay may be considered as the original place of the raw materials of the terracotta warriors and horses, the kiln sites might be located nearby the Qin’s Mausoleum.     

Thermoelectric power and carrier concentration of La2−xSrxCuO4(0.03 <x <0.17) single crystals

Thermoelectric power (S) measurements have been systematically investigated in high-quality La2-x-Srx-CuO4 single crystals with different Sr concentrations (x =0.03, 0.05, 0.07, 0.1, 0.12). In this doping range studied, S remains positive, and with the increase of Sr concentration,room-temperature thermoelectric power S (290 K) decreases.An empirical formula which presents a relationship between hole concentration P and room-temperature thermoelectric power S (290 K) is reported in this note. Based on the theory of metallic diffusion model, the experimental results are discussed.     

Modeling studies on the changes of the Subtropical High over the Western Pacific during the last 100 years

Intensity index series of Subtropical High over the Western Pacific was constructed according to modeling results by using of CCM3.6 atmospheric model with climatologically SST forcing for 30 years, and with observed SST forcing from 1900 to 2000. The index series constructed by modeling closely correlated to the observations for 1951-1999 (c.c. 0.72) and to the statistically reconstructed one for 1900-1950 (c.c. 0.51). The index series for 1900-1999 showed significant 80 a and 40 a periodicity. A minimum of persistence in seasonal variations of the index occurred in Autumn, which was called "Autumn Barrier". It may relate to the "Spring Barrier", found in the seasonal variations of SST over the Eastern Equatorial Pacific. The increasing trend shown in the index series for the 20th century may associate with the global warming caused by enhancement of greenhouse effect.     

Construction of multivalent DNA vaccines forMycobacterium tuberculosis and its immunogenicity

The coding regions of Ag85B MPT-64, and ESAT-6 secreted proteins were cloned initially into the eukaryotic expression vector pJW4303, then transformed to E. coli Top 10 strain for plasmid DNA extraction and further analysis. Plasmids containing the right insertion were sequenced to confirm their identity. COS7 cells were transfected with a mixture containing serially diluted plasmid DNA encoding three secreted proteins and Lipofectin (Gibco). The supernatants and pellets prepared from various cell lines were run on SDS-PAGE gel and the expression of these proteins in COS7 cells were demonstrated by immunoblot using polyclonal or monoclonal antiserum of M.TBH37Rv. 21 days after first vaccination of C57BL-6 mice by all three recombinant eukaryotic expressing vectors, antibody titer for Ag85B reached 1∶3200. 21 days after second vaccination, the antibody titer reached 1∶102400. The highest antibody levels induced by multivalent vaccines after the second injection were equal to or even greater than the highest antibody levels of single DNA vaccine reported in literature after third injections. Antibody titer of MPT-64 was 1∶50 after the first injection and it reached 1∶200 after the second injection. No antigen-specific antibody against ESAT-6 was detected in sera harvested from immunized mice 21 days after both injections. Antigen-specific IFN-g level of Ag85B was 110 pg/mL while no antigen-specific IFN- g level of ESAT-6 and MPT-64 was detected even after third injections. To our knowledge, it is the first time that studies of polyvalent recombinant DNA vaccines against TB were carried out in C57BL-6 mice. Our results indicated that multiple DNA vaccines could be used to enhance protective responses against M.TB.     

Minimum inductance optimal design for the gradient coil

In MRI (Magnetic Resonance Imaging), a crucial role of gradient coils is to image organism, meanwhile the inductance of the coils determines the speed of imaging. So it is of great importance to optimize designs of the gradient coils. The target field approach is an effective method to design the gradient coils. Having applied this method and performing many numerical tests, we achieved the designs of the x-、y-、z- gradient coils , with the linearity of the fields generated by the coils in a sphere of radius 0.30 m less than 5%, in which the inductance and resistance of the coils also meet the requirements.     

应用Elman神经网络的混沌时间序列预测

利用改进的 Elman神经网络对 3个典型的混沌时间序列在不同的噪声水平下进行预测 ,探讨了神经网络学习与泛化之间的关系 ,通过试凑法给出了 Elman最优的隐节点个数。并利用3种指标对预测结果进行了评估 ,结果显示 Elman网络对混沌时间序列预测的良好特性