Multi-receptive Field Aggregation Network for single image deraining

Image deraining is a significant problem that ensures the visual quality of images to prompt computer vision systems. However, due to the insufficiency of captured rain streaks features and global information, current image deraining methods often face the issues of rain streaks remaining and image blurring. In this paper, we propose a Multi-receptive Field Aggregation Network (MRFAN) to restore a cleaner rain-free image. Specifically, we construct a Multi-receptive Field Feature Extraction Block (MFEB) to capture rain features with different receptive fields. In MFEB, we design a Self-supervised Block (SSB) and an Aggregation Block (AGB). SSB can make the network adaptively focus on the critical rain features and rain-covered areas. AGB effectively aggregates and redistributes the multi-scale features to help the network simulate rain streaks better. Experiments show that our method achieves better results on both synthetic datasets and real-world rainy images.     

The optimal macro control strategies of service providers and micro service selection of users: quantification model based on synergetics

Multiple wireless technologies, possibly administered by same or different service providers (SPs), are expected to coexist in the rapidly-expanding heterogeneous networks. It is also economically beneficial for all SPs to cooperate with each other. In such model, one SP temporally shares a portion of its spectrum with other SPs at a certain price to allow better utilization of the available spectrum. To this end, users can make service selection decisions dynamically at micro level according to the performance satisfaction level and cost, depending on the pricing plans and spectrum sharing strategies of SPs at macro level. On the other hand, the users’ service selection population states in turn influence the control strategies of SPs. To model this dynamic interactive decision making problem, a novel approach based on synergetics is presented in this paper. The motion equations of population state decided by pricing and spectrum sharing are derived. Then, by taking into account the population state, an optimization problem is formulated to determine the optimal dynamic pricing and open access ratio for the SPs in order to maximize their respective benefits. A closed-loop feedback equilibrium is obtained as the solution of the formed optimization problem. Numerical results demonstrate the effectiveness and advantages of the proposed model in terms of dynamic control of the spectrum sharing and pricing as well as dynamic analysis of service selection.     

THEORY OF BAINITIC TRANSFORMATION: A MODEL WITH COUPLED FIRST-ORDER PHASE TRANSFORMATIONS

1.IntroductionBainite,discoveredinsteelmorethansixdecadesago[1],nowhasbecomeanenlargedtermofphysicalmetallurgyandmateriasscience.Itcharacterisesaseriesofphase-transformationproductsformedintherangeofintermediatetemperaturenotonlyinsteelbutalsoinothernonferrousalloysystemsandeveninsomeoxideceramics[2l.Beingofgreatpracticalandtheoreticalimportance,ithasbeenextensivelyinvestigatedanddeeplyconsideredI3.However,someessentialaspectsarestillhighlycontroversialI4].Theexistingtheorlesofthetwoschools,…     

低能离子束原位研究系统的建立及其应用

介绍一套具有质量分辨能力的低能离子束原位研究系统及其性能。在该系统中 ,低能离子束能量可调范围为 5～ 2 0 0eV ,束流为 1～ 6 μA ;在 10 0eV的能量下 ,束流密度可达 10 0 μA/cm2 ;离子束能量降至 10eV时 ,束流密度为 10～ 2 0 μA/cm2 。能量分散为± 0 5eV。离子的质量和能量分别由四极质谱仪和筒镜能量分析器表征 ,离子束束流和束流密度分布用两个法拉第杯进行测量。WienFilter的质量分辨率为 4 0。系统上装有XPS ,UPS ,LEED等分析仪器 ,确保在UHV条件下实现原位表面研究     

未知光源位置环境中物体形状恢复的神经网络方法研究

用神经网络方法解决未知光源位置环境中物体三维形状恢复的问题.对漫反射表 面,用神经网络方法由已知表面形状物体及其对应图像的灰度值进行学习,所得权值可视为 环境光源参数.由此可恢复同样光源环境中其它物体的三维形状.实验证明,神经网络方法可 以解决未知光源位置环境(包括多个光源)中漫反射表面物体的三维形状恢复问题.     

吉林火车站屋盖结构风荷载特性的试验研究

通过对吉林火车站的刚性模型风洞试验研究,得到在有无周边建筑群干扰下屋盖表面的平均风压系数、脉动风压系数以及屋盖的升力系数分布特性,同时对屋盖局部体型系数及结构风振响应进行详细的对比分析研究。研究结果表明:火车站屋盖结构表面主要呈现负风压(风吸力),屋盖合力升力系数均为负值。主站楼在迎风区屋檐、悬挑区域及屋面凸起的天窗位置气流分离较大,风压变化明显,出现较大的负风压系数;由于站台雨篷四周开敞,气流流经站台雨篷时较为顺畅,气流分离较小,因而站台雨篷风压系数较小。周边的干扰建筑整体上对屋面全风向局部体型系数有一定的遮挡效应,但也不能忽略部分区域增大的情况;周边建筑群对站台雨篷角部、主站楼凸起的天窗及东西两侧悬挑屋面区域风振响应影响较大。     

复合活化处理镍钛合金及磷灰石层的仿生沉积

利用H2O2和NaOH溶液对近等原子比NiTi记忆合金（SMA）进行复合活化处理,并采用XRD,SEM,XPS,FTIR等研究了活化处理后NiTi SMA的表面结构及其在模拟体液（SBF）中磷灰石涂层的仿生沉积过程.结果表明：复合活化处理后NiTi SMA表面为贫Ni、富含Ti-OH的蠕虫状活性层,由结晶度较低的TiO2,钛酸纳（Na2TiO3）及少量的Ni2O3组成,在SBF中表面活性层诱导磷灰石形核及生长,短期内在镍钛基体表面获得了理想的磷灰石层.     

晶粒细化对Cr-25Nb合金高温氧化行为的影响

采用电弧熔铸和机械合金化+热压烧结技术制备晶粒尺寸相差较大的Cr-25Nb合金,研究其在950及1200 ℃空气中的氧化行为。结果表明,熔铸态及机械合金化Cr-25Nb合金氧化后均没有发生Cr的单一外氧化,而形成了以Cr2O3为外层、NbCrO4为内层的双层氧化膜结构;机械合金化Cr-25Nb合金在950及1200 ℃的氧化速度均小于熔铸态合金,特别是在1200 ℃氧化100 h后,熔铸态Cr-25Nb合金的氧化增重是机械合金化合金的2倍多。这主要是因为晶粒细化促进了氧化膜内应力的释放,提高了氧化膜与基体的粘附性     

端羧酸丁腈橡胶环氧基酯增韧剂的合成与应用

采用端羧酸丁腈橡胶(CTBN)来合成1种新型的CTBN环氧基酯增韧剂。介绍了该增韧剂的合成工艺路线和条件。并通过应用实例阐述了该增韧剂在环氧固化体系中的应用效果。结果表明:加入一定量的催化剂可有效地提高酯化反应的速率;该增韧剂对环氧树脂固化体系有显著增韧效果。     

无填充块木结构的剪力墙变形实验研究

对于无填充块木结构的剪力墙,因为没有填充块来连接墙骨,其变形主要由墙骨变形产生.用曲线描述墙骨的变形,即可从墙顶的位移得知墙骨上每一点的变形.墙骨上的变形,主要由钉节点的变形引起,通过钉节点变形性能公式,可以得到墙骨变形与力的关系,叠加可得整片墙体的受力与变形关系.通过以上对力学模型和实际变形相关关系的分析,以及力学关系的推导,提出无填充块剪力墙的受力变形计算公式.将计算公式与实验结果相比较,二者吻合情况较好,表明提出的无填充块剪力墙的受力变形计算公式具有一定的实用价值.