Main trends on the simulation of the drying of refractory castables - Review

Refractory castables develop microstructures after curing that behave as partially saturated porous media. Upon heating (during its drying stage), the steam generated by the physical and chemically bond water can result in pore pressurization and explosive spalling. Numerical modeling can provide guidelines for designing safer heat-up profiles and also a better understanding of the mechanisms that lead to catastrophic damage. This work aims to review the fundamentals and models available, providing insightful thoughts on the current trends of the drying phenomena of ceramic compositions. The review also highlights that there are models better oriented to result in reasonable predictions of pore pressure values and others focused on a more accurate representation of the main physical phenomena that take place during heating. According to the findings, there are still various challenges to attain accurate models with high applicability capable of yielding safer and more efficient drying of refractory castables.     

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。     

基于LBSN动态异构网络的时间感知兴趣点推荐

随着基于位置社交网络(Location-Based Social Network,LBSN)的快速发展,兴趣点(Point-Of-Interest,POI)推荐可以帮助人们发现有趣的并吸引人的位置。针对签到数据的稀疏性和用户兴趣的动态性等挑战性问题,提出了基于LBSN动态异构网络的时间感知兴趣点推荐算法。在LBSN异构网络模式中增加会话节点类型。通过动态元路径,在用户和兴趣点语义关系之间有效地融入时间信息、位置信息和社交信息等。设置了用户-兴趣点之间的动态元路径集,并提出了动态路径实例的偏好度计算方法。采用矩阵分解模型对不同动态偏好矩阵进行矩阵分解。根据不同动态元路径的用户特征矩阵和兴趣点特征矩阵,获取用户在目标时间访问兴趣点的推荐列表。实验结果表明,与其他兴趣点推荐方法相比,所提方法在兴趣点推荐精确度上取得了较好的推荐结果,具有良好的应用前景。     

密文策略属性加密中的撤销控制方案

在云环境下使用数据共享功能时，由于云环境的复杂性，需要对数据进行安全保护和访问控制，这就要求使用加密机制。基于密文策略属性的加密（CP-ABE）是当前广泛使用的加密机制，它可以根据用户的属性来设置访问权限，任何具有合格访问权限的用户都可以访问数据。然而云是一个动态环境，有时可能只允许具有访问权限用户中的一部分用户访问数据，这就需要用户权限的撤销机制。然而，在CP-ABE中，访问权撤销或用户撤销是一个冗长且代价高昂的事件。所提出方案根据对CP-ABE流程的改进，在原密文中嵌入了可灵活控制的用户个人秘密，使得用户权限撤销时既不要求使用新访问策略的用户撤销数据，也不要求对数据进行重新加密，大幅减少撤销时的计算成本。与知名CP-ABE撤销方案对比，所提出方案的计算成本更低且具有良好的安全性。     

基于改进的Random Subspace的客户投诉分类方法

电信业的客户投诉不断增多而又亟待高效处理。针对电信客户投诉数据的特点，提出了一种面向高维数据的改进的集成学习分类方法。该方法综合考虑客户投诉中的文本信息及客户通讯状态信息，基于Random Subspace方法，以支持向量机（Support Vector Machine，SVM）为基分类器，采用证据推理（Evidential Reasoning，ER）规则为一种新的集成策略，构造分类模型对电信客户投诉进行分类。所提模型和方法在某电信公司客户投诉数据上进行了验证，实验结果显示该方法能够显著提高客户投诉分类的准确率和投诉处理效率。     

适用于无线体域网的动态口令认证协议

无线体域网中传输的是与生命高度相关的敏感数据，身份认证是信息安全保护的第一道防线。现有的基于人体生物信息的身份认证方案存在信息难提取、偶然性大和误差性大的问题，基于传统密码学的认证方案需较大计算资源和能量消耗，并不适用于无线体域网环境。为此，在动态口令和非对称加密机制基础上，提出一种适用于无线体域网的动态口令双向认证轻量协议，并对其进行形式化分析。通过理论证明、SVO逻辑推理及SPIN模型检测得出：该协议满足双向认证，且能够抵御重放攻击、伪装攻击、拒绝服务器攻击和口令离线攻击，具有较高安全性。     

基于碳纳米管的超级电容器聚吡咯电极复合方法

For the defect that the mechanical properties of polypyrrole supercapacitors decrease with charge and discharge cycles, and the cycle stability is poor, a compound method of polypyrrole electrode for supercapacitors based on multi-walled carbon nanotubes is proposed. The structure of the formed polypyrrole coated multi-walled carbon nanotubes effectively increase the specific surface area of the electrode material, the utilization rate of the active material and the electrical conductivity, improve the rapid charge and discharge performance of the electrode material, and greatly improve cyclic stability of polypyrrole. The composite electrode materi- al of polypyrrole and functionalized multi-walled carbon nanotubes for supercapacitors is prepared by pulse current deposition meth- od. It is scanned at a scanning rate of 1000mV?s -1 in a 3mol?L -1 potassium chloride electrolyte, after 100,000-cycle, the capacity on- ly declines by 16%.     

Stochastic collocation-based nonlinear analysis of concrete bridges with uncertain parameters

This paper focuses on the stochastic response of concrete bridges considering uncertainty in bearing and abutment stiffness. A multi-span simply supported bridge with concrete girders is selected. A 3D-dimensional model is prepared, and nonlinear response history analyses are performed. For the numerical dynamic simulation, the non-sampling stochastic method based on generalized polynomial chaos (gPC) expansion is utilised. The uncertain parameters include the vertical and shear stiffness of bearings and the lateral stiffness of abutments are presented by the truncated gPC expansions. Furthermore, the system response such as base shear, acceleration, velocity and displacement in different columns is presented by gPC expansion with unknown deterministic coefficients. The stochastic Galerkin projection is employed to calculate a set of deterministic equations. A non-intrusive solution, as a set of collocation points, determines the unknown gPC coefficients of the system response and the results are compared with Monte Carlo simulations. The key advantage of spectral discretization is the combination of the mentioned method with the spatial discretization, e.g. finite element model. This study also emphasises the accuracy in results and time efficiency of the proposed non-sampling method for uncertainty quantification of stochastic systems comparing to sampling procedure (e.g. Monte Carlo simulation).     

基于Faster R-CNN的颜色导向火焰检测

基于深度特征的目标检测方法Faster R-CNN在火焰检测任务上存在检测效率低的问题,因此提出了基于颜色引导的抛锚策略。该策略设计火焰颜色模型来限制锚的生成,即利用火焰颜色约束锚的生成区域,从而减少了初始锚的数量,提升了计算效率。为了进一步提高网络的计算效率,将区域生成网络中的卷积层替换成掩膜卷积。为了验证所提方法的检测效果,采用BoWFire和Corsician数据集进行验证。实验结果表明,该方法实际检测速度相较于原Faster R-CNN提高了10.1%,BoWFire上该方法的火焰检测F值为0.87,Corsician上该方法的准确度可达99.33%。所提方法可以提高火焰检测的效率,并能够准确检测图像中的火焰。     

基于优势学习的深度Q网络

强化学习问题中，同一状态下不同动作所对应的状态-动作值存在差距过小的现象，Q-Learning算法采用MAX进行动作选择时会出现过估计问题，且结合了Q-Learning的深度Q网络（Deep Q Net）同样存在过估计问题。为了缓解深度Q网络中存在的过估计问题，提出一种基于优势学习的深度Q网络，通过优势学习的方法构造一个更正项，利用目标值网络对更正项进行建模，同时与深度Q网络的评估函数进行求和作为新的评估函数。当选择的动作是最优动作时，更正项为零，不对评估函数的值进行改动，当选择的动作不是最优动作时，更正项的值为负，降低了非最优动作的评估值。和传统的深度Q网络相比，基于优势学习的深度Q网络在Playing Atari 2600的控制问题breakout、seaquest、phoenix、amidar中取得了更高的平均奖赏值，在krull、seaquest中取得了更加稳定的策略。