论传统文化中色彩元素与视觉传达的关系

传统文化包含众多的元素,作者对传统文化元素中的色彩元素在现代视觉传达设计中的运用与影响进行了研究,分别就中国传统色彩在包装设计、标志设计以及环境设计中的运用进行了论述,并阐释了如何实现现代元素与传统文化元素在现代视觉传达设计中的完美统一,最终达到在现代视觉传达设计中传承和发展优秀传统文化的目的.     

长春市某高校间歇采暖节能潜力分析

针对学校建筑采暖负荷特点,采用能耗分析的方法分析研究采用间歇供暖的节能性.两种运行方式耗热量对比,建筑面积为10.6 W㎡的学校建筑,采用间歇采暖运行方式可以节省能量2 913 734 kWh,相当于节约标准煤556t,并减少水泵耗电31 691 kWh,相当于节省原来用能的11.7%.由于采用间歇采暖的运行方式使用小流量供暖,不但节约了燃料消耗量,减少了污染物排放,而且也显著降低了运行费用,是一种较为理想的节能方法.     

Nd8-xDyxFe82Co6B4纳米复合永磁材料的磁性能

针对NdFeB纳米复合永磁材料的矫顽力偏低的现状,采用熔体快淬法制备了Nd8DyxFe82Co6B4（x=0,0．4,0．8,1．2,1．6）纳米复合永磁材料,研究了Dy和Co的添加对NdFeB纳米复合永磁材料的磁性能的影响．研究结果表明,Dy的添加可提高材料的各向异性场,提高矫顽力,但是Dy元素的添加使材料的剩磁降低,结晶温度升高,因此添加量不宜过多;复合添加Co原子不仅可以提高材料的剩磁和矫顽力,而且可以弥补只添加Dy元素使材料的结晶温度升高的缺陷．在最佳的热处理条件下,Dy含量x=0．8时的样品表现出最佳的磁性能,矫顽力jHc=524kA／m,剩磁Br=1．11T,最大磁能积（BH）max=158kJ／m^3．     

基于CRF的中文评论有效性挖掘产品特征

方面级意见挖掘的任务通常包括从客户评论中抽取产品的特征、与产品特征相关联的观点词识别以及观点的极性判断三个方面。围绕如何实现中文评论的方面级意见挖掘问题,提出了利用条件随机场实现中文评论的方面级意见挖掘的四个主要步骤：数据预处理、训练集准备、为条件随机场模型定义学习函数、应用模型标注新的评论数据。在此基础上,通过以五种实际产品的中文评论语料为数据集,对该方法进行了数据实验。实验结果表明,该方法针对不同类型观点元素的抽取在评估性能指标上大部分达到或超过80%。为了进一步验证所提出方法的有效性,将研究结果进行了差异显著性检验。结果显示,用CRF对中文评论进行方面级意见挖掘和对英文评论的方面意见挖掘的性能差异不大。最后,比较了三种不同方法的方面抽取精度和情感分类精度,实验结果表明,CRF方法优于词典化的隐马尔可夫模型和关联规则挖掘方法。     

Trace element geochemistry of the Obed Mountain deposit coals, Alberta, Canada

This study reports on the elemental concentrations and vertical variation of coal seams from the Obed Mountain deposit, Alberta Foothills, Canada. Results from two sections of Seam 1 show that the major elements (i.e. Al, Fe, Mg, K, Na, Ti, and Si) have high concentrations in intervals having high ash content, with the only exception of Ca. Similarities are apparent, in both sections, in the vertical variations of Th, U, Se, and Zn; Rb, Cs, and K; Sb, Mo, and W; Mn and Sr; and Ba, Cr, Co, Hf, and Sc. These similarities are also evident among the REEs, notably between Ce and La; also between Dy, Eu, and Sm. Most elements, with the exception of Ba and Sr are slightly more concentrated in Section 2 of Seam 1, located approximately 1.5Fig. km away from Section 1. Compared to Seam 1, Seam 2 has lower mean concentrations of elements. Boron in the coal ranges from 27 to 100 ppmw, though most values are less than 50 ppmw. Boron concentrations suggest a freshwater depositional environment. The element is depleted in the sedimentary partings (12–29 ppmw only) and is enriched in the coal interval near the roof and immediately beneath the partings. This enrichment shows possible downward mobilization of boron. Vertical variations of elements are helpful in delineating the boundaries between coal and sedimentary partings in the succession. The Obed Mountain coals are “clean” by world standards and their elemental concentrations are comparable with those in coals of a lower rank from the same coal formation used for power generation in Alberta. All sedimentary partings have low concentrations of Ba, Hf, Sc, Sr, Ta, Th, U, and REEs; this, along with the absence of an Eu negative anomaly suggests a non-volcanic origin for the partings.     

Selective separation of yttrium from rare earth elements by ion interaction chromatography. Analytical and larger scale applications

ABSTRACT

New ion interaction chromatographic (IIC) method employing Kromasil 100 C₁₈ column and tetra-n-butylammonium hydroxide (TBAOH) as an ion interaction reagent (IIR) is able to separate yttrium from Rare Earth Elements (REE) as anionic complexes with nitrilotriacetic acid (NTA). New method for the chromatographic determination of Y in REE mixture was devised and validated by the analysis of the certified reference material (CRM). Potential possibilities of the new chromatographic system for larger scale applications including macro-micro events were demonstrated.     

Trace element loads in the Great Lakes Basin: A reconnaissance

Water quality data for trace elements in the Great Lakes are relatively scarce, complicating the assessment of current trace element baselines and their distribution patterns. Here, we present concentration data for >40 major and trace elements in >100 samples from the Great Lakes connecting channels, surface waters, precipitation and select Canadian tributaries, to establish a high-level assessment of loading rates across the basin. Contrasting upstream-to-downstream trends were observed for the investigated trace elements, ranging from net-decreasing (>5-fold for e.g., Co, Tl, Y) to net-increasing surface water concentrations (>2-fold for e.g., Sb, U, As). Calculated loading rates reveal different, element-specific controls of runoff, connecting channel loads or precipitation on trace element occurrence. Lake-wide elemental mass-balances could be reasonably closed for conservative trace elements (e.g., Li, <53% residual) but not for others (e.g., rare earth elements with up to 5-fold discrepancies), reflective of general data scarcity and uncertainty in loading rates. In line with major water quality trends, spatial distribution patterns in Lakes Erie and Ontario display subtle near-shore to off-shore heterogeneity for a few trace elements (<1 order-of-magnitude for V or Se), but higher variability for trace elements with significant inputs derived from tributaries. This work provides important quantitative baseline data for trace elements in the Great Lakes that may help optimize surveillance and management strategies for the preservation of Great Lakes water quality.     

西北地区农业干旱对气象干旱的时空多角度响应

采用标准化降水蒸散指数和标准化土壤水指数分别表征气象干旱和农业干旱,基于干旱事件三维识别和气象干旱与农业干旱事件对的匹配结果,利用Copula函数条件概率分布和贝叶斯网络概率模型分析了西北地区农业干旱对气象干旱的时空多角度响应关系。结果表明：基于线性、非线性关系模型及Copula函数条件概率分布确立了西北地区农业干旱对气象干旱的干旱历时、干旱烈度、干旱面积及干旱迁移距离间最优响应关系分别为二次多项式模型、基于Frank Copula函数的非线性关系模型、指数模型和线性模型;建立的气象、农业干旱变量间的响应概率特征曲线可定量预测西北地区不同程度的气象干旱发展为农业干旱的概率;已知气象干旱特征条件下,根据最优响应关系及响应概率特征曲线可对农业干旱特征及其发生概率进行可靠预测。     

数学形态学在数字图像处理中的应用研究

传统边缘检测算法由于对噪声敏感,难以准确提取图像边缘,导致图像处理效果不佳.基于传统数学形态学算法中结构算子的方向性和尺寸几何的基础上进行算法改进.针对抗噪型碰撞腐蚀形态学边缘检测算子结构元素特征,采用不同大小结构元素组合来提取边缘特征,有效保证了图像细节的同时去掉较大噪声点.根据结构元素的方向性,利用同向结构元素图像的匹配来检测各边缘信息,确保不同向边缘信息的完整度.通过比较文本改进算法与传统的边缘检测算法对图像边缘检测效果表明:本文提出的改进算子在处理较大图像边缘检测时具有更快的检测速度,且图像边缘光滑,细节清晰,具备了更强的抗噪性能.     

基于互信息F统计量特征选择技术的地基气象云图分类

在地基气象云图的云状(云类)识别研究中,基于局部二值模式(Local Binary Pattern,LBP)描述子的特征选择技术由于它的简单性和有效性成为最通用的方法.然而,LBP特征的高维特性使得云状识别的性能和计算开销不能令人满意.为此,本文提出一种基于互信息构造的F检验统计量的LBP特征选择算法,可以实现高维LBP特征的有效降维,同时保证云状识别的准确性,极大减少了特征选择过程的计算开销.