Simulations of NBI fast ion loss in the presence of toroidal field ripple on EAST

NBI fast ion losses in the presence of the toroidal field ripple on EAST have been investigated by using the orbit code GYCAVA and the NBI code TGCO. The ripple effect was included in the upgraded version of the GYCAVA code. It is found that loss regions of NBI fast ions are mainly on the low field side near the edge in the presence of ripple. For co-current NBIs, the synergy effect of ripple and Coulomb collision on fast ion losses is dominant, and fast trapped ions located on the low field side are easily lost. The ripple well loss and the ripple stochastic loss of fast ions have been identified from the heat loads of co-current NBI fast ions. The ripple stochastic loss and the collisioninduced loss are much larger than the ripple well loss. Heat loads of lost fast ions are mainly localized on the right side of the radio frequency wave antennas from the inside view toward the first wall. For counter-current NBIs, the first orbit loss due to the magnetic drift is the dominant loss channel. In addition, fast ion loss fraction with ripple and collision for each NBI linearly increases with the effective charge number, which is related to the pitch angle scattering effect.     

抗菌聚酯切片中锌含量测定方法的研究

介绍了一种以氧化锌为添加剂的抗菌聚酯,分别从样品处理、试验仪器、实验条件等方面讲述了几种测定抗菌聚酯切片中锌含量的测定方法,对是否含二氧化钛的抗菌聚酯切片给出建议测定方法,并总结了几种分析方法的优缺点.     

胶莱盆地南部大盛群、王氏群的划分与对比

胶莱盆地在胶东半岛地质构造及造山带研究中占有重要地位,其巨厚的白垩系陆相沉积是石油系统多年的研究目标。胶莱盆地南部白垩纪的岩石地层序列较其北部更为复杂,其认识几经反复,其中尤以大盛群与王氏群和莱阳群的部分单位的关系为甚。本次工作通过详细对比其沉积层序、相互关系、古生物组合、岩石学特征及同位素年龄等因素,认为大盛群与王氏群属同一沉积层序,沉积基底相同、沉积环境相似、古生物群落时代一致,二者具有同时代、同沉积相、同岩石组合的一致性,应是一套地层,建议今后使用统一地层名称。     

Characteristics of reaction and product microstructure during light calcination of magnesite in transport bed

A gas-heating laboratory transport bed was adopted to simulate the industrial transportbed reactor for flash calcination of magnesite, and a method based on TG analysis of a reacted sample was further developed to calculate the decomposition rate or conversion of its containing MgCO₃. The study investigated how the conversion and product reactivity as well as microstructure vary with reaction conditions including temperature, particle size and times of re-calcination for powder magnesite. Magnesite powder (<150 μm) calcination is a quick reaction that reaches 98% decomposition of its containing MgCO₃ in 1—2 s,corroborating the feasibility of magnesite flash-calcination in transport bed reactors. The coloration time given by the citric acid chromogenic method was 17—55 s and 294 s for the obtained products from transport and fixed beds, respectively. This proves the obviously higher activity and thus improved microstructure of the product from transport bed. During the calcination process, the MgO grain size of the product gradually increases, and the surface structure changes from loose and porous to dense and smooth. This structural change can be completed within a few seconds.     

北京山区推理公式设计洪水计算若干问题研究

北京山区推理公式是北京地区无资料山区小流域设计洪水计算的基本方法之一.结合北京山区原推理公式与改进后推理公式存在问题,分析了利用推理公式计算山区小流域设计洪水的适用性和可靠性,并提出相关使用建议.针对当前推理公式方法计算设计洪峰流量与洪水过程线计算中采用的降雨和径流资料不一致问题,提出将山区推理公式与瞬时单位线方法组合应用,解决无资料小流域设计洪水过程线计算的基本思路和方法.研究表明,该方法理论上可行,方法更为合理,建议推广使用.     

改进的CAGAN在虚拟试衣中的应用

针对CAGAN（Conditional Analogy GAN）换衣后效果模糊，在目标衣服与原始衣服长短不一致时效果一般，相对目标衣服保留过少的细节等问题做了相关研究并对CAGAN进行了改进，提出了新的虚拟试衣方式。经过改进的CAGAN生成一个粗糙的结果，由该结果得到目标衣服穿在模特身上改变形状后的mask，接下来利用mask对目标衣服进行变形，综合变形的衣服和第一步的结果便得到最终的试衣图像。实验结果表明，该方法解决了前面存在的问题，而且取得了非常好的效果。     

混菌矿化增强再生粗骨料的物理力学性能

为克服单一微生物培养成本高且矿化鲁棒性不足的缺陷，提出了一种混菌矿化增强再生粗骨料物理力学性能的方法.通过筛选矿化效率较高的好氧嗜碱混菌，考察了混菌矿化对再生粗骨料物理力学性能和混凝土抗压强度的影响.结果表明：相同增强时间下，混菌比纯菌呈现出更优异的矿化增强效果；随着混菌矿化增强时间的延长，再生粗骨料吸水率和压碎指标呈现出先减小后增大的趋势，最优增强时间为15 d；采用矿化增强再生粗骨料制备的再生混凝土抗压强度提高幅度达到22.1%.     

基于形态特征提取的急性下壁心肌梗死BiLSTM网络辅助诊断算法

急性下壁心肌梗死是一种病发急、进展快、致死率高的心脏疾病,该文提出一种新颖的基于形态特征提取的BiLSTM神经网络分类的急性下壁心肌梗死辅助诊断算法,可大幅度提高医生对急性下壁心肌梗死疾病的诊断效率并有助于及时确诊.算法包括:对胸痛中心数据库心拍信号进行降噪及心拍分割;根据临床心内科医学诊断指南提取了12导联波形距离特征和分导联波形幅值特征;依据提取的特征搭建LSTM与BiLSTM神经网络进行心拍的分类识别;使用PTB公开数据库和胸痛中心数据库多临床中心进行交叉验证.实验结果表明,加入胸痛中心真实临床数据后,基于形态特征提取BiLSTM神经网络的急性下壁心肌梗死辅助诊断算法准确率达到99.72％,精度达到99.53％,灵敏度达到100.00％,同时F1-Score达到99.76.该算法比其他现有算法准确率提高至少1％,该项研究具有非常重要的临床应用价值.