基于快速神经网络架构搜索的鲁棒图像水印网络算法

为解决深度学习在图像水印算法中计算量大且模型冗余的问题，提高图像水印算法在抵抗噪声、旋转和剪裁等攻击时的鲁棒性，提出基于快速神经网络架构搜索（neural architecture search，NAS）的鲁棒图像水印网络算法。通过多项式分布学习快速神经网络架构搜索算法，在预设的搜索空间中搜索最优网络结构，进行图像水印的高效嵌入与鲁棒提取。首先，将子网络中线性连接的全卷积层设置为独立的神经单元结构，并参数化表示结构单元内节点的连接，预先设定结构单元内每个神经元操作的搜索空间；其次，在完成一个批次的数据集训练后，依据神经元操作中的被采样次数和平均损失函数值动态更新概率；最后，重新训练搜索完成的网络。水印网络模型的参数量较原始网络模型缩减了92%以上，大大缩短了模型训练时间。由于搜索得到的网络结构更为紧凑，本文算法具有较高的时间性能和较好的实验效果，在隐藏图像时，对空域信息的依赖比原始网络更少。对改进前后的2个网络进行了大量鲁棒性实验，对比发现，本文算法在CIFAR-10数据集上对抵抗椒盐噪声和旋转、移除像素行（列）等攻击优势显著；在ImageNet数据集上对抵抗椒盐高斯噪声、旋转、中值滤波、高斯滤波、JPEG压缩、裁剪等攻击优势显著，特别是对随机移除行（列）和椒盐噪声有较强的鲁棒性。     

Excellent UV-absorbing cotton fabric with high durability and hydrophobicity based on surface-initiated ATRP of polymerizable organic UV-absorber

Cellulose - Ultraviolet (UV) protective cotton fabric is a convenient and reliable way to protect human body from sunlight. Herein, a polymerizable UV-absorber,...     

基于不确定控制理论的最优策略研究

本文在不确定理论的框架下,研究一类带背景状态变量的最优控制模型.在乐观值准则下,利用不确定动态规划的方法,证明了不确定最优性原则,得到最优性方程.作为应用,求解一个固定缴费(DC)型养老金的最优投资策略问题,在乐观值准则下,以工资变量为背景状态变量,建立养老金模型.通过求解不确定最优性方程得到最优投资策略和最优支付率.     

Understanding the solubility and crystallization of molybdenum in high-sodium borosilicate glasses: effect of lanthanum oxide

Journal of Radioanalytical and Nuclear Chemistry - The simplified 8-oxide system (SiO2-B2O3-Na2O-Al2O3-CaO-TiO2-MoO3-La2O3) has been prepared under two types of cooling methods, and analyzed with...     

ON THE EXISTENCE WITH EXPONENTIAL DECAY AND THE BLOW-UP OF SOLUTIONS FOR COUPLED SYSTEMS OF SEMI-LINEAR CORNER-DEGENERATE PARABOLIC EQUATIONS WITH SINGULAR POTENTIALS

In this article,we study the initial boundary value problem of coupled semi-linear degenerate parabolic equations with a singular potential term on manifolds with corner singularities.Firstly,we introduce the corner type weighted p-Sobolev spaces and the weighted corner type Sobolev inequality,the Poincare′inequality,and the Hardy inequality.Then,by using the potential well method and the inequality mentioned above,we obtain an existence theorem of global solutions with exponential decay and show the blow-up in finite time of solutions for both cases with low initial energy and critical initial energy.Significantly,the relation between the above two phenomena is derived as a sharp condition.Moreover,we show that the global existence also holds for the case of a potential well family.     

烧结温度对煤系高岭土制备堇青石陶瓷的影响

我国的煤系高岭土具有明显的资源优势,应对其进行充分地研究和合理地应用,以缓和优质高岭土资源枯竭的燃眉之急.以鄂尔多斯煤系高岭土、鄂尔多斯滑石、工业氧化铝为原料,通过反应烧结一步合成堇青石,探讨了堇青石合成温度范围,以拓宽该煤系高岭土的应用范围.通过高温显微镜观察堇青石试样的合成温度范围为1400～1420℃.经1420℃合成的堇青石试样的吸水率为16.11;,显气孔率为28.44;,体积密度为1.77 g/cm3,径向烧成收缩为1.77;,厚度烧成收缩为4.80;,抗折强度为43.56 MPa.镁铝尖晶石是合成堇青石的重要中间产物,适当提高烧结温度可以促进镁铝尖晶石相向堇青石相转变,有利于堇青石晶粒的生长与发育.烧结温度的变化是堇青石晶粒发生变化的原因,同时导致液相量的变化,最终影响堇青石试样显微结构和烧结性能.     

A Shell-by-Shell Approach for Synthesis of Mesoporous Multi-Shelled Hollow MOFs for Catalytic Applications

Over the past two decades, advanced materials with hollow interiors have received significant attention in materials research owing to their great application potential across a vast number of technological fields. Though with great difficulty, multi-shelled hollow metal–organic frameworks (MSHMs) have also been successfully synthesized in recent years. Herein, a rational shell-by-shell soft-templating protocol has been devised to fabricate highly uniform multi-shelled hollow cobalt-imidazole-based MOF (ZIF-67). For the first time, it has become possible to endow mesoporosity to this new type of functional material (i.e., mesoporous MOFs). When used as carrier materials in catalytic reactions, in principle, these mesoporous MSHMs with high surface area not only improve the dispersity of metal nanoparticles (NPs), but also efficiently facilitate the mass diffusion of the reactions, resulting in enhanced catalyst activity. Moreover, the obtained MSHMs/M nanocomposites serve as base-metal bifunctional catalysts for one-pot oxidation-Knoevenagel condensation cascade reaction, in which the MSHMs itself serves as a pristine active catalyst in addition to its role of catalyst support. The results demonstrate that excellent multifunctional catalysts can be achieved via preparing intrinsically microporous bulk MOFs into extrinsically mesoporous MSHMs which possess many structural merits that conventional bulk MOFs do not have.     

强度解调型光纤光栅温度传感实验

光纤光栅在工程上应用广泛,但由于存在解调系统复杂、成本高,尤其是需要使用光纤光谱仪等波长解调仪器,使得光纤光栅很难走入本科教学的实验课堂.本文提出了一种基于窄线宽DFB激光器的强度解调方案,极大地简化了光纤光栅传感器的解调系统,完全满足较高分辨率和实时检测的系统要求.合理安排的温度传感实验可以非常直观地展示光纤光栅的线性传感性能,使得工程化的光纤光栅传感技术轻松走进本科教学课堂,具有极高的推广和应用潜力.