基于高阶图卷积自编码器的网络流量预测

网络流量预测是有效保障用户QoS措施之一。当前深度学习为基础的网络算法预测中没有充分利用网络拓扑信息。为此,提出了基于高阶图卷积自编码器的网络流量预测模型。该流量预测模型基于软件定义网络（SDN）架构,利用高阶图卷积网络（GCN）获取网络拓扑中的多跳邻域之间的流量相互影响关系,采用门控递归单元（GRU）获取网络的时间相关性信息,利用自编码模型来实现无监督学习和预测。在Abilene网络上采用真实数据进行了仿真对比分析试验,结果表明,提出的方法在网络流量检测方面的MAPE值为41.56%,低于其它深度学习的方法,同时预测准确率方面也达到最优。     

铝合金单层球面网壳结构阻尼特性

通过现场实测对铝合金板式节点单层球面网壳结构的阻尼比进行了分析研究.针对一平面尺寸45 m×45 m,矢高2.86 m的铝合金板式节点单层球面网壳结构,设计了11种工况,通过现场实测采集了160条人工激励下的节点加速度自由衰减振动信号以及6条环境激励下的节点加速度振动信号,采用解析模态分解法(AMD)结合希尔伯特变换识别结构的自振频率和阻尼比.对所得数据进行分析,建议铝合金板式节点单层球面网壳的结构阻尼比取4%.运用实测阻尼比数据建立有限元模型分析结构动力响应,对应实测节点加速度响应曲线和有限元计算得到的响应曲线吻合较好,所测得阻尼值可为现行规范修订提供依据,为结构动力分析与工程设计提供参考.     

Mechanical,electrical, and thermal properties of the directionally solidified Bi-Zn-Al ternary eutectic alloy

A Bi-2.0Zn-0.2Al (wt%) ternary eutectic alloy was prepared using a vacuum melting furnace and a casting furnace. The samples were directionally solidified upwards at a constant growth rate (V = 18.4 μm/s) under different temperature gradients (G = 1.15–3.44 K/mm) and at a constant temperature gradient (G = 2.66 K/mm) under different growth rates (V = 8.3–500 μm/s) in a Bridgman-type directional solidification furnace. The dependence of microstructure parameter (λ) on the solidification parameters (G and V) and that of the microhardness (Hv) on the microstructure and solidification parameters were investigated. The resistivity (ρ) measurements of the studied alloy were performed using the standard four-point-probe method, and the temperature coefficient of resistivity (α) was calculated from the ρ-T curve. The enthalpy (ΔH) and the specific heat (C_p) values were determined by differential scanning calorimetry analysis. In addition, the thermal conductivities of samples, obtained using the Wiedemann-Franz and Smith-Palmer equations, were compared with the experimental results. The results revealed that, the thermal conductivity values obtained using the Wiedemann-Franz and Smith-Palmer equations for the Bi-2.0Zn-0.2Al (wt%) alloy are in the range of 5.2–6.5 W/Km and 15.2–16.4 W/Km, respectively.     

Tool life modeling for evaluating the effects of cutting speed and reinforcements on the machining of particle reinforced metal matrix composites

The wear of cutting tools in the machining of 2024Al alloy composites reinforced with Al₂O₃ particles using varying sizes and volume fractions of particles up to 23.3vol% was investigated by a turning process using coated carbide tools K10 and TP30 at different cutting speeds. Machining tests were performed with a plan of experiments based on the Taguchi method. The tool life model was developed in terms of cutting speed, size, and volume fraction of particles by multiple linear regressions. The analysis of variance (ANOVA) was also employed to carry out the effects of these parameters on the cutting tool life. The test results show that the tool life decreases with the increase of cutting speed for both cutting tools K10 and TP30, and the tool life of the K10 tool is significantly longer than that of the TP30 tool. For the tool life, cutting speed is found to be the most effective factor followed by particle content and particle size, respectively. The predicted tool life of cutting tools is found to be in very good agreement with the experimentally observed ones.     

Comparative stability analyses of traditional and selective room-and-pillar mining techniques for sub-horizontal tungsten veins

The stability and productivity concerning a modification on the traditional room and pillar for a new selective technique at the Portuguese Panasqueira Mine have been described. The traditional room-and-pillar stoping uses 5.0-m wide rooms with 3.0 m×3.0 m pillars, while the selective room-and-pillar mining technique consists in stoping with rooms of 4.0 m wide and pillars of 4 m×4 m with a subsequent selective cutting of the quartz veins at the mid pillar of approximately 0.5 m high, to obtain a pillar section with an area of 3.0 m×3.0 m. The stability and productivity analyses indicate that the selective technique obtains smaller average pillar safety factor, more rock mass displacement, more extraction and selectivity ratios, and ore grade improvement, compared with the traditional technique. These results show that the selective technique is also more convenient. This proposed selective room-and-pillar mining technique is applicable to any sub-horizontal narrow quartz veins with wolfram, gold, etc. such as the famous La Rinconada gold mine in the Peruvian Andes.     

钨合金精密磨削砂轮磨损及对表面质量的影响机制

为了满足钨合金零件的高精度和高完整性表面加工需求,实现钨合金磨削质量的控制和加工工艺的优化,通过分析砂轮磨粒种类和结合剂类型对砂轮磨损形式的关系,明确了超硬磨料砂轮在钨合金磨削加工中的优势.研究了磨粒粒度和磨削参数对钨合金磨削加工质量的影响规律,为制定合理的磨削工艺提供依据.通过磨削试验获得钨合金高精度和高完整性表面磨削加工工艺.研究结果表明,钨合金磨削过程中砂轮易磨损,超硬磨料砂轮更适合钨合金的磨削加工.金属结合剂金刚石砂轮在钨合金磨削加工的表面质量和加工精度方面表现出优越性.通过改进磨削参数,钨合金精密磨削后表面粗糙度可达18.9 nm,实现了镜面磨削效果.     

基于区间解析冗余关系的参数不确定性研究

本文主要目的是拓展基于区间解析冗余关系法I-ARRs(Interval valued Analytical Redundancy Relations)对不确定性混杂系统的故障检测。该方法产生一组对故障估计的阈值,一旦标称残差超过阈值,则故障能被检测。现有的故障检测方法主要是针对参数性故障,因此对测量信号导数的估计呈现出各种限制。而且在I-ARRs中对测量信号导数十分敏感,这不可避免地涉及到对残差的估计。为解决这一问题,利用鲁棒滑模微分器方法得到对含噪声测量信号导数最优估计,降低残差噪声,以减小虚警、漏报,从而提高鲁棒性。通过典型电力电子电路BUCK电路仿真,验证了此方法的有效性。     

An aqueous extract of Liu Wei Di Huang Wan alters the labeling of blood constituents with Technetium-99m

The Liu Wei Di Huang Wan is a formula of a traditional Chinese medicine that is used to treat asthma patients and has been shown to have several important properties,such as antioxidant and free radical scavenging activities.The influence of an extract of Liu Wei Di Huang Wan on the labeling of blood constituents with technetium-99m was investigated.Anticoagulated blood(Wistar rats) was incubated with the extract,stannous chloride and technetium-99m,as sodium pertechnetate.Samples were centrifuged and aliquots of plasma and blood cells were separated and precipitated with trichloroacetic acid,to obtain soluble and insoluble fractions of the blood constituents.The percentage of radioactivity(%ATI) in all the fractions was determined.The analysis of the results shows that the extract at the highest concentration used(70 mg/mL) decreased significantly(P<0.05) the %ATI(from 96.48 ± 1.19 to 54.46 ± 7.38) on blood cells compartment,(from 81.11 ± 4.15 to 61.33 ± 4.74) on insoluble fractions of blood cells and(from 65.91 ± 2.44 to 13.15 ± 3.62) on insoluble fractions of plasma.In conclusion,the results suggest that the substances present on this extract can alter this labeling process,probably due to(i) redox properties(antioxidant and chelator activities) and/or(ii) specific actions in the binding sites where the 99mTc would be bound on the blood constituents.As a consequence,precaution is suggested on the interpretation of the nuclear medicine results from performed with blood constituents labeled with 99m Tc in patients that have undertaken LWDHW,although the current findings were obtained in experimental animal models.     

基于双重自注意力机制的人脸图像修复

人脸图像修复旨在修复输入人脸图像中的缺失区域,生成令人满意的高质量修复结果.然而当存在大面积缺失时,直接修复缺失人脸图像十分困难,此时修复网络的全局上下文信息感知能力是影响修复结果的关键.鉴于此,本文提出了软硬注意力相结合的双重自注意力模块.该模块通过全局相似度计算来获得软硬两种注意力特征,之后对两种注意力特征进行自适应融合,进而提高修复网络对全局上下文信息的感知能力.此外,本文进一步提出了多尺度生成对抗网络以加强对修复结果的监督,促使修复网络生成更高质量的修复结果.实验结果表明,本文方法在定量和定性评测上均优于五种先进的对比方法.     

夏侃紫外线

在炎炎烈日的夏天,你依然想拥有润泽白皙的肌肤吗?那么有关紫外线的知识、有关防晒护肤的新概念可是不可不知的。 何谓紫外线? 由太阳放射出来的光线,以波长来分类,大致可分为伽马线、X光、紫外线、可视光和红外线五种。紫外线是自太阳射至地球表面一种眼睛看不到的光线。而紫外线又可再细分为长波长紫外线(UVA)、中波长紫外线(UVB)和短波长紫