Computational algorithms for simulating the grain structure formed on steel billets using cellular automaton and chaos theories

The development of some computational algorithms based on cellular automaton was described to simulate the structures formed during the solidification of steel products. The algorithms described take results from the steel thermal behavior and heat removal previously calculated using a simulator developed by present authors in a previous work. Stored time is used for displaying the steel transition from liquid to mushy and solid. And it is also used to command computational subroutines that reproduce nucleation and grain growth. These routines are logically programmed using the programming language C++ and are based on a simultaneous solution of numerical methods (stochastic and deterministic) to create a graphical representation of different grain structures formed. The grain structure obtained is displayed on the computer screen using a graphical user interface (GUI). The chaos theory and random generation numbers are included in the algorithms to simulate the heterogeneity of grain sizes and morphologies.     

Effect of the formation process of transient liquid phase (TLP) on the interface structure of TiAl joints

TiAl has been joined employing the transient liquid phase (TLP) bonding with Ti combined with Cu, Ni or Fe foils. Experimental results showed that though the interface structures of the joints are quite different, all the joined zones are composed of five sublayers, i.e. two diffusion zones, two interfacial zones and an interlayer. It has been convinced that the formation process of the transient liquid phase controls the diffusion behavior of melting point depressant (MPD) Cu, Ni, and Fe atoms, which leads to form different interface structures of the joints.     

Action mechanisms of a new erythrocyte-derived depressing factor

To investigate the action mechanisms of a new erythrocyte-derived depressing factor (EDDF), the focus is placed on the effect of EDDF on both cytosolic and nuclear free calcium (Ca²⁺) transportation in vascular smooth muscle cell (VSMC), as well as the apoptosis and cell cycle of VSMC of rats. EDDF has been extracted from human erythrocytes. The changes of Ca²⁺ levels in cytoplasm ([Ca²⁺]_i) and nucleus ([Ca²⁺]_n) have been observed using a laser scanning confocal microscope together with fluo-3/AM as a calcium indicator. Flow cytometric technique was used to study the effect of EDDF on cell cycle and apoptosis of VSMC. [Ca²⁺], and [Ca²⁺]_n were significantly decreased through several different pathways: ( i ) it reduced the Ca²⁺ influx by blocking L-type voltage-dependent calcium channel (L-VDC) and R-type voltage-dependent calcium channel (R-VDC); (ii) it inhibited the Ca²⁺ release from inositol 1, 4, 5-trisphosphate (IP₃) sensitive calcium store; and (iii) activated Ca²⁺-ATPase of sarcoplasmic reticulum (SR) and promoted the transportation of Ca²⁺ from cytoplasm to SR. However, EDDF seemed to have little inhibitory effect on the Ca²⁺ release from ryonodine sensitive calcium pool. It was also found that EDDF (10⁻⁴ g/mL) significantly decreased the proportion of S phase of human umbilical vein (HUV) and inhibited the proliferation of VSMC induced by angiotensin II (Angll, 10⁻⁵ mol/L). The apopotosis did not occur when VSMC was cultured under normal condition. While VSMC apoptosis was induced by Angll (10⁻⁵ mol/L) and EDDF (10⁻⁴ g/mL) seemed to have little effect on it. The inhibitory effect of EDDF on the elevation of [Ca²⁺]_i and [Ca²⁺]_n of VSMC might play an essential role in its action mechanisms and the ways it affects the Ca²⁺ handling of VSMC demonstrate that EDDF was different from other endogenous blood pressure regulators and some known antihypertensive drugs. EDDF could inhibit the proliferation of VSMC, which indicated that it might be beneficial to the prevention and treatment of hypertension and arteriosclerosis.     

以改进的Tietze法合成对位取代苯酚氧苷类化合物

对Tietze报道的合成苯酚氧苷的方法进行了改进,以全乙酰糖代替1-O-三甲基硅烷基乙酰糖作反应的原料,以无水SnCl4代替三氟甲磺酸三甲基硅酯作反应的促进剂,缩短了反应路线,降低了工艺成本.利用改进的方法设计合成了三个对位取代苯酚氧苷类化合物1-O-(4-硝基苯基)-2,3,5-三-O-乙酰基-β-D-呋喃核糖苷、1-O-(4-乙酰基苯基)-2,3,5-三-O-乙酰基-β-D-呋喃核糖苷和1-O-(4-苯甲酰氧基苯基)-2,3,5-三-O-乙酰基-β-D-呋喃核糖苷.考察了原料比例、促进剂用量和溶剂对反应的影响:反应的促进剂无水SnCl4用量需大于化学计量;溶剂为干燥的二氯甲烷,以非干燥的二氯甲烷或乙腈为溶剂的反应则会有副产物4-硝基苯基乙酸酯的产生,副产物的出现可能为乙酰基参与的糖苷键断裂所致.     

Synthesis and characterization of boron-doped NiO thin films produced by spray pyrolysis

Boron-doped NiO thin films were prepared on glass substrates at 400℃ by airbrush spraying method using a solution of nickel nitrate hexahydrate. Their physical properties were investigated as a function of dopant concentration. From X-ray diffraction patterns, it is observed that the films have cubic structure with lattice parameters varying with boron concentration. The morphologies of the films were examined by using scanning electron microscopy, and the grain sizes were measured to be around 30–50 nm. Optical measurements show that the band gap energies of the films first decrease then increase with increasing boron concentration. The resistivities of the films were determined by four point probe method, and the changes in resistivity with boron concentration were investigated.     

鄱阳湖生态经济区A级旅游景区空间结构研究

通过ArcGIS和地理集中系数、基尼系数、洛伦兹曲线等数理方法,对鄱阳湖生态经济区A级旅游景区的空间结构进行了探讨和分析,将2004、2007、2010和2012年4个时间段A级景区的数量进行对比,并将新增景区与交通干线、重要水系进行叠置,动态分析 A 级景区空间结构的演化规律.研究结果表明:鄱阳湖生态经济区 A 级旅游景区在空间分布类型上属于集群分布,空间结构上呈多核-边缘结构,等级规模上为纺锤型结构,空间上呈现非均衡的分布态势,时间上呈现沿交通干线和主要水系的演变规律.     

利巴韦林衍生物的合成研究

以2’,3’-异丙叉利巴韦林(Ⅰ)为原料,通过碱催化与3个不同的酰氯缩合酯化和酸催化水解脱除异丙叉保护基两步反应合成了3个化合物,即1-(5-对硝基苯甲酰基-β-D-呋喃核糖基)-1,2,4-三氮唑-3-甲酰胺(Ⅲa),1-[5-(3,4,5-三苄氧基苯甲酰基)-β-D-呋喃核糖基]-1,2,4-三氮唑-3-甲酰胺(Ⅲb)和1-(5-苯甲酰基-β-D-呋喃核糖基)-1,2,4-三氮唑-3-甲酰胺(Ⅲc),Ⅲa和Ⅲb通过催化氢化得到1-(5-对氨基苯甲酰基-β-D-呋喃核糖基)-1,2,4-三氮唑-3-甲酰胺(Ⅳa)和1-[5-(3,4,5-三羟基苯甲酰基)-β-D-呋喃核糖基]-1,2,4-三氮唑-3-甲酰胺(Ⅳb),考察了不同碱和不同溶剂对酯化反应的影响,不同酸对脱除异丙叉保护基的影响以及催化剂Pd/C的用量对催化氢化的影响.中间体及目标产物均通过IR,1HNMR进行了结构表征.     

基于多分类器动态集成的电信客户流失预测

本文提出了一种新的基于多分类器动态选择与优化集成的电信客户流失预测集成模型.首先使用K均值聚类算法对训练集样本进行分区;然后分别使用Naive-Bayes算法、多层感知机算法和J48算法构建各分区客户流失预测子分类器;最后对各分区子分类器进行线性集成,并使用人工蜂群算法优化其集成权重.当测试样本由聚类算法判断出其归属区域后,再分别使用分区子分类器进行预测,最后使用优化权重进行线性集成.实验结果表明:动态集成模型优于单模型;基于人工蜂群算法优化集成模型优于其它集成模型.     

基于元数据的电网通信资源数据校核方法

由于缺乏有效管理手段,导致电网通信资源数据的完整性、一致性、有效性无法保障,影响电网通信专业服务质量。为此,提出了基于元数据的电网通信资源数据校核方法。该方法分为三个过程：利用网络捕包得到电网通信资源数据,将捕获的数据存入磁盘列阵内和电网通信资源数据多层次校核。多层次校核内的信息层校核将XSD推荐标准作为元数据实施驱动,利用XSD校核引擎实现信息校核;模型层校核将元数据模型内同公共信息模型OWL描述的本体文件作为元数据,校核电网通信网络各系统资源信息内提取的RDF格式模型实例;数据层校核以元数据模型内拓扑规则元数据作为校核依据,校核电网各系统中常见的孤岛、内外环等拓扑结构问题。实验结果表明：所提方法能有效验证电网通信资源数据的一致性、有效性和完整性,且校核精度较高达95%以上。     

细粒度全局感知多聚焦图像融合网络

多聚焦图像融合是图像融合的一个重要分支,在显微成像等方面具有广泛的应用.针对多聚焦融合中存在的纹理细节不清晰、聚焦区域误判等问题,本文从空间及通道信息全局关注的角度出发,结合Swin Transformer中的移动窗口自注意力机制和深度可分离卷积设计了一个全局信息编码-解码网络,采用综合损失函数进行图像重构任务的无监督学习;从特征邻域信息重要性的角度出发,引入了改进的拉普拉斯能量和函数在特征域进行图像聚焦属性的判别,增强图像聚焦区域判别的细粒度效果.与7种经典图像融合算法比较,本文算法在定性和定量分析中均取得了先进的融合性能表现,对原始图像的聚焦区域信息具有更高的保真效果.