Microstructure and mechanical properties of α/β-Si3N4 composite ceramics with novel ternary additives prepared via spark plasma sintering

In this study, α/β-Si₃N₄ composite ceramics with high hardness and toughness were fabricated by adopting two different novel ternary additives, ZrN–AlN–Al₂O₃/Y₂O₃, and spark plasma sintering at 1550 °C under 40 MPa. The phase composition, microstructure, grain distribution, crack propagation process and mechanical properties of sintered bulk were investigated. Results demonstrated that the sintered α/β-Si₃N₄ composite ceramics with ZrN–AlN–Al₂O₃ contained the most α phase, which resulted in a maximum Vickers hardness of 18.41 ± 0.31 GPa. In the α/β-Si₃N₄ composite ceramics with ZrN–AlN–Y₂O₃ additives, Zr₃AlN MAX-phase and ZrO phase were found and their formation mechanisms were explained. The fracture appearance presented coarser elongated β-Si₃N₄ grains and denser microstructure when 20 wt% TiC particles were mixed into Si₃N₄ matrix, meanwhile, exhibited maximum mean grain diameter of 0.98 ± 0.24 μm. As a result, the compact α/β-Si₃N₄ composite ceramics containing ZrN–AlN–Y₂O₃ additives and TiC particles displayed the optimal bending strength and fracture toughness of 822.63 ± 28.75 MPa and 8.53 ± 0.21 MPa?m^1/2, respectively. Moreover, the synergistic toughening of rod-like β-Si₃N₄ grains and TiC reinforced particles revealed the beneficial effect on the enhanced fracture toughness of Si₃N₄ ceramic matrix.     

基于粒子群优化和支持向量机的花粉浓度预测模型

为了提高花粉浓度预报的准确率，解决现有花粉浓度预报准确率不高的问题，提出了一种基于粒子群优化（PSO）算法和支持向量机（SVM）的花粉浓度预报模型。首先，综合考虑气温、气温日较差、相对湿度、降水量、风力、日照时数等多种气象要素，选择与花粉浓度相关性较强的气象要素构成特征向量；其次，利用特征向量与花粉浓度数据建立SVM预测模型，并使用PSO算法找出最优参数；然后利用最优参数优化花粉浓度预测模型；最后，使用优化后的模型对花粉未来24 h浓度进行预测，并与未优化的SVM、多元线性回归法（MLR）、反向神经网络（BPNN）作对比。此外使用优化后的模型对某市南郊观象台和密云两个站点进行逐日花粉浓度预测。实验结果表明，相比其他预报方法，所提方法能有效提高花粉浓度未来24 h预测精度，并具有较高的泛化能力。     

计及Spark和属性权重的售电套餐推荐方法

针对电力市场用户群庞大，交易过程中售电套餐选择困难的问题，在Spark环境下设计了一种售电套餐推荐方法，同时也解决了售电套餐推荐过程中在大数据环境下的可扩展性及实时性问题。首先，计算出每个套餐属性的权重值，从而计算得到售电套餐综合相似度。然后，计及用户和套餐两方面提出一种售电套餐推荐方法，实现售电套餐的精准推荐。实验表明，提出的推荐方法能够明显提高推荐的准确度，并且在分布式环境下具有良好的推荐效率和可扩展性。     

Preparation of nanocrystalline nickel oxide from nickel hydroxide using spark plasma sintering and inverse Hall-Petch related densification

Nanocrystalline nickel oxide (NiO) was prepared from nickel hydroxide by Spark plasma sintering (SPS) and the mechanisms involved in the densification of NiO were studied. Reverse precipitated nickel hydroxide powders were SPS processed at 400, 600 and 700?°C with 70?MPa pressure. Pure NiO with 12?nm crystallite size formed after 400?°C sintering process. However NiO grains had grown to 18 and 38?nm after 600 and 700?°C sintering respectively. NiO pellets prepared using 600 and 700?°C SPS sintering schedules had relative densities of 83% and 94% respectively. Two displacement rate regimes were observed during densification of NiO in both 600 and 700?°C sintering processes. Decomposition of nickel hydroxide and particle sliding of NiO led to first displacement rate maximum while inverse Hall-Petch based plastic deformation facilitated densification during the constant second displacement rate regime. No densification occurred during sintering holding times indicating the limited role that diffusion played during densification.     

一种基于Spark的不确定数据集频繁模式挖掘算法

如何在海量不确定数据集中提高频繁模式挖掘性能是目前研究的热点.传统算法大多是以期望、概率或者权重等单一指标为数据项集支持度，在大数据背景下，同时考虑概率和权重支持度的算法难以兼顾其执行效率.为此，本文提出一种基于Spark的不确定数据集频繁模式挖掘算法（UWEFP），首先，为了同时兼顾数据项的概率和权重，计算一项集的最大概率权重值并进行剪枝；然后，为了减少对数据集的多次扫描，结合Spark框架的优点，设计了一种具有FP-tree特征的新颖的UWEFP-tree结构进行模式树的构建及挖掘；最后在Spark环境下，以UCI数据集进行实验验证.实验结果表明本文的方法在保证挖掘结果的同时，提高了效率.     

多孔硅（PS）制备新方法

介绍了两种制备多孔硅的方法：电火花刻蚀法和激光辐射腐蚀法。讨论了这两种新方法制备的多孔硅样品的结构和发光特性，同时，与电化学法制备的多孔硅的结构和发光特性进行了比较。最后指出这两种新方法对于多孔硅形成机理和发光机制研究是有所帮助的。     

Granulation of a hardmetal powder for spark plasma sintering

A fluidized bed granulation method, pressure swing granulation (PSG), was applied to granulation of a hardmetal powder without pressing lubricants for making the upstream process of spark plasma sintering (SPS) more efficient.

The properties of the granules were examined and compared with those of spray dried granules and extruded ones under the present system using a sieve.

Spherical granules between 0.15 and 0.84 mm in diameter difficult to obtain by the spray drying were obtained with high yield. The flowability of the granules was far better than that of spray dried granules and similar to that of extruded ones. Iron contamination and oxidization during pressure swing granulation were tolerable to the real production.     

基于流量摘要的僵尸网络检测

随着僵尸网络的日益进化,检测和防范僵尸网络攻击成为网络安全研究的重要任务.现有的研究很少考虑到僵尸网络中的时序模式,并且在实时僵尸网络检测中效果不佳,也无法检测未知的僵尸网络.针对这些问题,本文提出了基于流量摘要的僵尸网络检测方法,首先将原始流数据按照源主机地址聚合,划分适当的时间窗口生成流量摘要记录,然后构建决策树、随机森林和XGBoost机器学习分类模型.在CTU-13数据集上的实验结果表明,本文提出的方法能够有效检测僵尸流量,并且能够检测未知僵尸网络,此外,借助Spark技术也能满足现实应用中快速检测的需要.     

基于Apache Spark的地震观测数据噪声功率谱计算

为解决单机环境下海量地震观测数据计算和分析效率低下的问题,提出一种基于分布式架构的地震观测数据的存储、计算和分析处理方法,选择噪声功率谱复杂计算过程的应用场景进行实现.基于Hadoop在海量数据处理上的性能优势,在分布式文件存储系统HDFS上进行地震观测数据的存储和调度,研究测震数据噪声功率谱的质量评估方法在Spark分布式计算架构上的实现,采用弹性数据集Spark RDD将计算任务自动分配到计算节点,解析存储在HDFS中的测震波形数据,计算结果采用RowKey方式放入分布式数据库HBase中,实现了长周期地震噪声功率谱结果的存储和提取.计算结果表明,基于Spark分布式架构的该方法可以支撑TB级海量数据的处理,并且具有较高的处理效率,可应用于海量地震观测数据的分析计算.     

一种基于深度强化学习的Spark Streaming参数优化方法

Spark Streaming作为主流的开源分布式流分析框架,性能优化是目前的研究热点之一。在Spark Streaming性能优化中,业务场景下的配置参数优化是其性能提升的重要因素。在Spark Streaming系统中,可配置的参数有200多个,对参数调优人员的经验要求较高,未经优化的参数配置会影响流作业执行性能。因此,针对Spark Streaming的参数配置优化问题,提出一种基于深度强化学习的Spark Streaming参数优化方法（DQN-SSPO）,将Spark Streaming参数优化配置问题转化为深度强化学习模型训练中的最大回报获得问题,并提出权重状态空间转移方法来增加模型训练获得高反馈奖励的概率。在3种典型的流分析任务上进行实验,结果表明经参数优化后Spark Streaming上的流作业性能在总调度时间上平均缩减27.93%,在总处理时间上平均缩减42%。